09 - Scientific and electric apparatus and instruments
41 - Education, entertainment, sporting and cultural services
42 - Scientific, technological and industrial services, research and design
44 - Medical, veterinary, hygienic and cosmetic services; agriculture, horticulture and forestry services
Goods & Services
Downloadable computer software for managing information
regarding tracking, compliance, and motivation with a
health, fitness, and nutrition program; downloadable
software for providing personal training services, coaching,
workouts, nutrition, fitness, health, fertility, and sleep
assessments and feedback; downloadable software using
artificial intelligence (AI) for providing personal training
services, coaching, workouts, nutrition, fitness, health,
fertility, and sleep assessments and feedback; downloadable
software for creating personalized fitness training,
nutrition, sleep, and well-being programs. Providing information, counseling, and advice relating to
fitness, fitness training, and activity; physical fitness
conditioning classes; fitness boot camps; yoga classes;
providing online non-downloadable videos in the fields of
health, wellness, fitness, exercise, and nutrition;
providing and conducting classes, seminars and workshops in
the fields of health, wellness, fitness, exercise, and
nutrition; online journals, namely, blogs featuring
commentary, advice and information in the fields of health,
wellness, sleep, fitness and nutrition; providing coaching
services in the fields of diet, nutrition, wellness,
fitness, health, wellness, mental health, and disease and
condition management; personal coaching services in the
nature of offering calls or chats, notifications, the
ability to track activities, incentive management solutions
and wellness challenges; physical fitness consultation
services. Providing online non-downloadable computer software managing
information regarding tracking, compliance, and motivation
with a health, fitness, and nutrition program; providing
online non-downloadable software for providing personal
training services, coaching, workouts, nutrition, fitness,
health, fertility, and sleep assessments and feedback;
providing online non-downloadable software using artificial
intelligence (AI) for providing personal training services,
coaching, workouts, nutrition, fitness, health, fertility,
and sleep assessments and feedback; providing online
non-downloadable software for creating personalized fitness
training, nutrition, sleep, and well-being programs. Providing information, counseling, and advice relating to
health, nutrition, fertility, body fat, body mass index,
sleep, stress, blood oxygen levels, and heart rate;
healthcare services; healthcare services, namely, services
to enable effective management of one or more chronic
conditions; healthcare services, namely, wellness and
prevention programs, healthcare management programs, disease
management programs and medical condition management
programs, and chronic care management; providing wellness
services, namely, personal assessments, personalized
routines, maintenance schedules, and counseling; providing
healthcare and wellness services, namely, personal
assessments, personalized routines, maintenance schedules,
fitness evaluations, and counseling; providing personal
lifestyle wellness evaluation and consultation; providing
educational information about healthcare for wellness
program coordinators; consulting services in the field of
health.
09 - Scientific and electric apparatus and instruments
41 - Education, entertainment, sporting and cultural services
42 - Scientific, technological and industrial services, research and design
44 - Medical, veterinary, hygienic and cosmetic services; agriculture, horticulture and forestry services
Goods & Services
Downloadable computer software for managing information
regarding tracking, compliance, and motivation with a
health, fitness, and nutrition program; downloadable
software for providing personal training services, coaching,
workouts, nutrition, fitness, health, fertility, and sleep
assessments and feedback; downloadable software using
artificial intelligence (AI) for providing personal training
services, coaching, workouts, nutrition, fitness, health,
fertility, and sleep assessments and feedback; downloadable
software for creating personalized fitness training,
nutrition, sleep, and well-being programs; downloadable
computer software for receiving, processing, transmitting,
tracking, and displaying information relating to health,
fitness, nutrition, fertility, activity, body fat, body mass
index, sleep, stress, blood oxygen levels, and heart rate;
downloadable computer software using artificial intelligence
(AI) for receiving, processing, transmitting, tracking, and
displaying information relating to health, fitness,
nutrition, fertility, activity, body fat, body mass index,
sleep, stress, blood oxygen levels, and heart rate. Providing information, counseling, and advice relating to
fitness, fitness training, and activity; physical fitness
conditioning classes; fitness boot camps; yoga classes;
providing online non-downloadable videos in the fields of
health, wellness, fitness, exercise, and nutrition;
providing and conducting classes, seminars and workshops in
the fields of health, wellness, fitness, exercise, and
nutrition; online journals, namely, blogs featuring
commentary, advice and information in the fields of health,
wellness, sleep, fitness and nutrition; providing coaching
services in the fields of diet, nutrition, wellness,
fitness, health, wellness, mental health, and disease and
condition management; personal coaching services in the
nature of offering calls or chats, notifications, the
ability to track activities, incentive management solutions
and wellness challenges; physical fitness consultation
services. Providing online non-downloadable software for providing
personal training services, coaching, workouts, nutrition,
fitness, health, fertility, and sleep assessments and
feedback; providing online non-downloadable software using
artificial intelligence (AI) for providing personal training
services, coaching, workouts, nutrition, fitness, health,
fertility, and sleep assessments and feedback; providing
online non-downloadable software for creating personalized
fitness training, nutrition, sleep, and well-being programs;
providing online non-downloadable computer software for
receiving, processing, transmitting, tracking, and
displaying information relating to health, fitness,
nutrition, fertility, activity, body fat, body mass index,
sleep, stress, blood oxygen levels, and heart rate;
providing online non-downloadable computer software using
artificial intelligence (AI) for receiving, processing,
transmitting, tracking, and displaying information relating
health, fitness, nutrition, fertility, activity, body fat,
body mass index, sleep, stress, blood oxygen levels, and
heart rate; providing online non-downloadable computer
software for managing information regarding tracking,
compliance, and motivation with a health, fitness, and
nutrition program. Providing information, counseling, and advice relating to
health, nutrition, fertility, body fat, body mass index,
sleep, stress, blood oxygen levels, and heart rate;
healthcare services; healthcare services, namely, services
to enable effective management of one or more chronic
conditions; healthcare services, namely, wellness and
prevention programs, healthcare management programs, disease
management programs and medical condition management
programs, and chronic care management; providing wellness
services, namely, personal assessments, personalized
routines, maintenance schedules, and counseling; providing
healthcare and wellness services, namely, personal
assessments, personalized routines, maintenance schedules,
fitness evaluations, and counseling; providing personal
lifestyle wellness evaluation and consultation; providing
educational information about healthcare for wellness
program coordinators; consulting services in the field of
health; corporate wellness services, namely, providing
assistance and consultation to corporate clients to help
their employees make health, fitness, wellness and
nutritional changes in their daily living to improve health
in the nature of wellness and health-related consulting
services.
Techniques and apparatuses are described for implementing multifaceted thermal-response shaping. In example aspects, a thermal control system manages heat that is generated by a subsystem of a system-on-chip to protect the subsystem from being damaged due to overheating, to maintain reliability of the system-on-chip, and to avoid creating a potentially unsafe situation for the user to operate a computing device with the system-on-chip. The thermal control system triggers the subsystem to operate at different operation points to shape a thermal response of the system-on-chip. The thermal control system appropriately shapes the thermal response based on an evaluation of two or more metrics, such as temperature and power, which can be associated with similar or different time scales. With this flexible, multifaceted approach, the thermal control system can determine an operation point that can improve the user experience and does not compromise a safety or reliability of the subsystem.
G05B 19/4155 - Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form characterised by programme execution, i.e. part programme or machine function execution, e.g. selection of a programme
The present document describes techniques for use in troubleshooting and debugging complex circuits, devices, and systems. These techniques may be implemented to provide hardware-assisted instruction-level debugging. By way of an example, techniques are provided that may be implemented in an apparatus having a plurality of state sequencers. The techniques may include receiving signals from a plurality of state sequencers, and storing an instruction history indicated, at least in part, by one or more of the signals received from a selected one of the state sequencers. The stored instruction history may indicate at least a state of the selected state sequencer during a specific cycle of instruction execution.
A method, a system, or one or more computer-readable storage media for training a machine learning model. The method includes receiving multiple training examples, each having a set of unmasked tokens for a first modality and a set of unmasked tokens for a second modality. For each training example, an encoder processes the sets of unmasked tokens for the first modality and the second modality to generate a set of embedded tokens for the first modality. A decoder processes at least the set of embedded tokens to reconstruct a set of masked tokens for the first modality. A loss is determined based on the reconstructed set of masked tokens and a set of masked tokens for the first modality according to an objective function, which is optimized to decrease the losses for the multiple training examples. The model parameters of the machine learning model are updated based on the optimization.
A computing device may receive inputted text and perform, using one or more neural networks, on-device grammar checking of a sequence of words in the inputted text, including determining, using the one or more neural networks, a grammatically correct version of the sequence of words and determining that the sequence of words does not match the grammatically correct version of the sequence of words. The computing device may, in response to determining that the sequence of words does not match the grammatically correct version of the sequence of words, output, for display at a display device, at least a portion of the grammatically correct version of the sequence of words as a suggested replacement for at least a sequence of the sequence of words in the inputted text.
A method includes receiving image data characterizing a face of an individual and receiving a voice-based command captured by a microphone of a wearable device worn by a user. The voice-based command includes a natural language command spoken by the user that requests the wearable device to identify the individual. The method also includes performing person identification to identify the individual by extracting an evaluation vector from the image data characterizing the face of the individual, determining the evaluation vector matches a reference vector for the individual, and determining an identify of the individual. The method also includes providing, for output from the wearable device, an identification cue that conveys an identity of the induvial to the user.
Encoding and decoding uses a multi-tree partitioning scheme comprising multiple categories (trees). Each category has a set of block split-type options such that each resultant sub-partition in each available block split-type option also belongs to one of the multiple categories. A category of the multi-tree recursive partitioning scheme is identified, and a block is partitioned using a split-type option belonging to the category of the multi-tree recursive partitioning scheme. The partitioning scheme is a recursive partitioning scheme, and the resulting sub-blocks may be encoded and subsequently decoded. Each category of the multi-tree recursive partitioning scheme may be defined by a unique, unordered pair (p, q) of block sizes p*2m×q*2n, p and q are odd, positive integers, and m and n are positive integers.
H04N 19/119 - Adaptive subdivision aspects e.g. subdivision of a picture into rectangular or non-rectangular coding blocks
H04N 19/176 - Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
A gantry crane for attachment to an immersion cooling system includes rails attachable to a tank of the immersion cooling system. The gantry also includes a collapsible frame configured to travel along the rails. A connector is adapted for connecting to a computer hardware component intended to be cooled in the immersion cooling system, and the gantry includes a winch mounted to the frame and configured to raise and lower the connector.
B66C 19/02 - Cranes comprising trolleys or crabs running on fixed or movable bridges or gantries collapsible
B66C 17/04 - Overhead travelling cranes comprising one or more substantially-horizontal girders the ends of which are directly supported by wheels or rollers running on tracks carried by spaced supports with lifting beams, e.g. slewable beams, carrying load-engaging elements, e.g. magnets, hooks
H05K 7/20 - Modifications to facilitate cooling, ventilating, or heating
A computer-implemented method for generating a unified machine learning model using a neural network on a data processing apparatus is described. The method includes the data processing apparatus determining respective learning targets for each of a plurality of object verticals. The data processing apparatus determines the respective learning targets based on two or more embedding outputs of the neural network. The method also includes the data processing apparatus training the neural network to identify data associated with each of the plurality of object verticals. The data processing apparatus trains the neural network using the respective learning targets and based on a first loss function. The data processing apparatus uses the neural network trained to generate a unified machine learning model, where the model is configured to identify particular data items associated with each of the plurality of object verticals.
G06V 10/44 - Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersectionsConnectivity analysis, e.g. of connected components
G06V 10/764 - Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
Improved methods are provided for generating, via a noise-diffusion iterative process, depth maps or optical flow maps from input images. Also provided are improved methods for training the machine learning model(s) employed in the iterative process and for augmenting die set of training data used to train such models. By translating the depth or optical flow map prediction process into the noise diffusion context, improved performance with respect to compute cost, training data, requirements, model size, and output quality are obtained. Additionally, the noise diffusion context allows models trained as described herein to generate maps de novo from target color images and/or to begin from initial ‘guess’ maps (e.g., noisy maps, maps containing holes) when generating improved output maps, natively incorporating the imperfect prior information represented by such initial maps.
In response to a user interacting with a tangible peripheral assistant control device (e.g., depressing a button of the device), causing an automated assistant to perform one or more actions. The action(s) performed can be based on input previously provided by the user in configuring the peripheral assistant control device. The action(s) performed in response to interaction with the peripheral assistant control device can vary based on one or more conditions, such as which user is currently active, where the peripheral assistant control device is currently located (which can optionally be inferred based on which of multiple assistant computing devices the button is paired with), and/or the current state of one or more smart devices and/or other devices (e.g., as determined based on a device topology). A utility of the peripheral assistant control device can be automatically extended beyond what was specifically requested by a user during configuration.
Systems and methods for providing a visual representation output descriptive of a preliminary diagnosis can include obtaining radiograph data, processing the radiograph data with a machine-learned model to generate one or more classification outputs, and generating the visual representation output based on the one or more classification outputs. The visual representation output can be generated such that the visual representation output can be provided for display on a plurality of different display types with a plurality of different technical capabilities.
A61B 6/46 - Arrangements for interfacing with the operator or the patient
A61B 6/50 - Apparatus or devices for radiation diagnosisApparatus or devices for radiation diagnosis combined with radiation therapy equipment specially adapted for specific body partsApparatus or devices for radiation diagnosisApparatus or devices for radiation diagnosis combined with radiation therapy equipment specially adapted for specific clinical applications
G06V 10/25 - Determination of region of interest [ROI] or a volume of interest [VOI]
G06V 10/26 - Segmentation of patterns in the image fieldCutting or merging of image elements to establish the pattern region, e.g. clustering-based techniquesDetection of occlusion
G06V 10/764 - Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
Systems and methods for unimodal visual search can include obtaining an image query, processing the image query to generate an intent determination, determining a particular prompt based on the intent determination, determining a plurality of search results, and processing the particular prompt, the image query and at least a subset of the plurality of search results to generate a model-generated response. The generative model can perform different response renderings based on the different prompts associated with different intents.
G06F 16/583 - Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for automated generation of animations. The system obtains digital content data comprising one or more digital components, obtains an animation template that defines a structure for the animation, processes input data comprising at least the digital content data to determine one or more style features of the animation, and generates animation data defining the animation by applying the determined style features to the one or more digital components and the animation template.
Techniques and systems are disclosed that perform content generation with integrated automated formatting using word processing applications that deploy large language models (LLMs). The techniques include receiving, a natural language (NL) query for a synthetic content for a document, identifying formatting rules of the document, and generating an augmented query that includes a representation of at least a portion of the NL query and a representation of the one or more formatting rules of the document. The techniques further include providing the augmented query to an LLM and updating the document with the synthetic content generated by the LLM in response to the augmented query.
An example method includes determining a switching frequency of a switched mode power supply that includes capacitors; and responsive to determining that the switching frequency of the switched mode power supply is within an audio hand, adjusting the switching frequency of the switched mode power supply to a different frequency within the audio band.
H02M 3/157 - Conversion of DC power input into DC power output without intermediate conversion into AC by static converters using discharge tubes with control electrode or semiconductor devices with control electrode using devices of a triode or transistor type requiring continuous application of a control signal using semiconductor devices only with automatic control of output voltage or current, e.g. switching regulators with digital control
H02M 3/156 - Conversion of DC power input into DC power output without intermediate conversion into AC by static converters using discharge tubes with control electrode or semiconductor devices with control electrode using devices of a triode or transistor type requiring continuous application of a control signal using semiconductor devices only with automatic control of output voltage or current, e.g. switching regulators
A method for providing an unobtrusive self-view to participants of a virtual conference is provided. The method includes receiving a self-view video stream including a self-view of a user participating in a virtual meeting. The video stream is being acquired by a camera of the client device. The method includes causing the self-view video stream to be presented in a first mode in a first self-view portion adjacent to a graphical user interface (GUI) control panel of a GUI displayed on the client device of the user. The GUI control panel can include a first control element to control the camera. The method can further include receiving a switch self-view command of the user, and responsive to receiving the switch self-view command, causing the self-view video stream to be presented in in a second mode in a second self-view portion located outside of the GUI control panel of the GUI displayed on the client device of the user.
Systems and methods for prompt tuning can utilize previously-learned prompts for the initialization of tuning for prompts on different tasks that may differ from the task associated with the previously-learned prompt. The prompt being utilized for initialization can be a generic prompt and/or may be a prompt selected based on a determined similarity between two or more task embeddings.
Provided are systems and methods which more efficiency train embedding models through the use of a cache of item embeddings for candidate items over a number of training iterations. The cached item embeddings can be “stale” embeddings that were generated by a previous version of the model at a previous training iteration. Specifically, at each iteration, the (potentially stale) item embeddings included in the cache can be used when generating similarity scores that are the basis for sampling a number of items to use as negatives in the current training iteration. For example, a Gumbel-Max sampling approach can be used to sample negative items that will enable an approximation of a true gradient. New embeddings can be generated for the sampled negative items and can be used to train the model at the current iteration.
G06F 12/0875 - Addressing of a memory level in which the access to the desired data or data block requires associative addressing means, e.g. caches with dedicated cache, e.g. instruction or stack
G06F 12/0891 - Addressing of a memory level in which the access to the desired data or data block requires associative addressing means, e.g. caches using clearing, invalidating or resetting means
Methods, systems, and apparatus, including computer programs encoded on computer-storage media, for time-limited key derivation. In some implementations, a module receives a request that provides a key identifier and obtain a control value indicating a future time. The module obtains a counter value indicating a first time, where the counter value is based on a current state of a counter, and where the counter is configured to update the state of the counter at a predetermined frequency. The module generates a comparison result based on comparing the control value indicating the future time with the counter value indicating the first time. The module generates a key based on the key identifier, the control value, the comparison result, and a stored random number, and the module provides the key in response to the request.
A computing device may receive, from a first software application executing at the computing device, a deep link and may determine a second application to handle the deep link. The Computing device may, in response to determining tire second software application to handle the deep link, determine one or more interactions with a user interface of the second software application to navigate the user interface of the second software application to specific content of the second software application indicated by the deep link. The computing device may perform, without user interaction, the one or more interactions with the user interface of the second software application to navigate the user interface of the second software application to the specific content of the second software application indicated by the deep link.
G06F 3/0481 - Interaction techniques based on graphical user interfaces [GUI] based on specific properties of the displayed interaction object or a metaphor-based environment, e.g. interaction with desktop elements like windows or icons, or assisted by a cursor's changing behaviour or appearance
G06F 3/0484 - Interaction techniques based on graphical user interfaces [GUI] for the control of specific functions or operations, e.g. selecting or manipulating an object, an image or a displayed text element, setting a parameter value or selecting a range
G06F 3/0488 - Interaction techniques based on graphical user interfaces [GUI] using specific features provided by the input device, e.g. functions controlled by the rotation of a mouse with dual sensing arrangements, or of the nature of the input device, e.g. tap gestures based on pressure sensed by a digitiser using a touch-screen or digitiser, e.g. input of commands through traced gestures
G06F 9/451 - Execution arrangements for user interfaces
G06F 16/955 - Retrieval from the web using information identifiers, e.g. uniform resource locators [URL]
23.
PROACTIVE PERSONALIZATION OF MULTIMEDIA CONTENT AND DIALOG CONTENT THROUGH UTILIZATION OF LARGE LANGUAGE MODEL(S)
Implementations described herein relate to causing a client device to initiate streaming of multimedia content and causing the client device to render dialog content before and/or during the streaming of the multimedia content. Processor(s) can generate a structured large language model (LLM) query that can be processed to generate LLM output. The LLM output can include, for example, an indication of the multimedia content and the dialog content. In some implementations, the processors(s) can determine when to generate an additional structured LLM query to continue the streaming of the multimedia content, and proactively cause additional LLM output to be generated based on the additional structured LLM query. In additional or alternative implementations, the processor(s) can determine when to cause the dialog content to be rendered with respect to the streaming of the multimedia content.
G06F 16/435 - Filtering based on additional data, e.g. user or group profiles
G10L 13/08 - Text analysis or generation of parameters for speech synthesis out of text, e.g. grapheme to phoneme translation, prosody generation or stress or intonation determination
24.
SESSION-BASED USER AWARENESS IN LARGE LANGUAGE MODELS
Implementations are described herein for using the information about user engagement with large language model (LLM) output as a subsequent input into an LLM, so that the LLM is able to provide, for rendition on one or more output devices, a subsequent output that is tailored to the user. In various implementations, based on one or more input device signals, a user engagement event with an element of a first LLM output generated using a LLM and rendered using one or more output devices may be detected. Additional information about the element of the first LLM output may be captured and used to generate at least part of a subsequent input prompt for the LLM. The subsequent input prompt may be processed using the LLM to generate a subsequent LLM output for rendition on one or more of the output devices.
An example folding device includes a hinge assembly; a first assembly rotatably connected to the hinge assembly about a first axis; a second assembly rotatably connected to the hinge assembly about a second axis; and a continuous display spanning the hinge assembly and connected to the first assembly and the second assembly, wherein the hinge assembly comprises: a synchronization slider configured to synchronize movement of the first assembly and the second assembly, wherein, as the first assembly and the second assembly rotate, the synchronization slider translates along a direction parallel to the first axis and the second axis, and wherein the synchronization slider includes one or more projections perpendicular to the direction; and a latching track disposed within the hinge assembly and configured to interface with the one or more projections of the synchronization slider to provide a force to retain the folding device in a closed state.
A wireless communication method and associated apparatus are configured to selectively initiate (506) a random access (RA) procedure based on whether the apparatus has an access stratum identifier. The apparatus communicates (502) with a reader device and receives a paging message associated with the RA procedure. The apparatus determines (504) the presence of the access stratum identifier and selectively initiates the RA procedure based on the determination.
There is disclosed a method for federated learning implemented by a client device. The method comprises obtaining a model input and processing, by a neural network comprising one or more neural network layers, the model input to generate a model output. The method further comprises caching the activations of one or more selected layers of the neural network generated from processing the model input by the neural network and generating a label based upon a user interaction with the model output. The method further comprises transmitting, to a server, training data based upon the cached activations and corresponding labels for generating one or more steering vectors for the one or more selected layers.
Size information associated with a sub-block partition of a chroma block is received from a compressed bitstream. Sub-block partition information indicating a sub-block partition position and a sub-block partition direction is derived for the chroma block based on previously coded information. A transform subblock is defined within a residual block of the chroma block using the size information and the sub-block partition information. An inverse transform is then applied to the transform subblock.
H04N 19/12 - Selection from among a plurality of transforms or standards, e.g. selection between discrete cosine transform [DCT] and sub-band transform or selection between H.263 and H.264
H04N 19/176 - Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
H04N 19/61 - Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding in combination with predictive coding
H04N 19/70 - Methods or arrangements for coding, decoding, compressing or decompressing digital video signals characterised by syntax aspects related to video coding, e.g. related to compression standards
29.
COORDINATING WORK ACROSS COMPUTE RESOURCES OF A WEARABLE DEVICE
An example method includes receiving, by a shared platform of a wearable device, a job from an application of the wearable device. The wearable device includes a set of compute resources including a microcontroller unit, the shared platform being hosted, at least in part, by the microcontroller unit. Further, the shared platform is configured to send and receive data from the set of compute resources via one or more platform abstraction application programming interfaces associated with each compute resource of the set of compute resources. The method also includes determining, by the shared platform, a target compute resource of the set of compute resources based on, at least, hardware topology information of the wearable device. The method additionally includes offloading, by the shared platform, the job to the target compute resource via a respective platform abstraction application programming interface associated with the target compute resource.
A method including in a first training process, receiving first 2D training images, generating a scene representation based on the first 2D training images using a machine learning model, generating a loss based on comparing the scene representation to a ground-truth scene representation, and training the machine learning model based on the loss, and in a second training process, receiving a first video frame including second 2D training images, receiving a second video frame including third 2D training images the first frame and the second frame being captured sequentially in time, generating a first intermediate prediction based on the first video frame using the machine learning model, generating a second intermediate prediction based on the second video frame using the machine learning model, generating a loss based on the first intermediate prediction and the second intermediate prediction, and training the machine learning model based on the loss.
The technology is directed to systems, methods, and computer-readable mediums for predicting, and in some cases, mitigating or otherwise addressing, effects human-initiated and non-human initated changes to a network may have. One or more temporal network graphs representative of the network may be constructed. One or more temporal network graphs representative of the network with a proposed change may be constructed. An effect on the network the proposed change will have may be predicted using a graph neural network (GNN). The predicted effect may be output.
Techniques of encoding a multi-channel audio signal include using a generative multi-channel audio synthesis and coding approach that describes each pseudosource in terms of a reference signal and spatial information. Whereas existing SA model coding methods are based on direct deterministic encoding of the SA model sequence, a stochastic method is used to generate the SA model sequence. The generation can be subject to conditioning to obtain a rendering that is perceptually identical to a particular original signal or to a signal class or rely solely on learned behavior. The method complements any SA-model conditioning information with knowledge learned in a training stage to facilitate a plausible spatial rendering.
The technology is generally directed to a method of mapping fiber networks. The fiber networks may include a plurality of cables, such as fiber optic cables. The cable may be divided into segments. Each cable has a first end segment and a second end segment, each with a known location. When there is a perturbation that causes the cable to vibrate, each segment of the cable may experience an associated strain at a different time. Based on the known location of the perturbation sources, the known location of the end segments, and the relative time that the perturbation is detected at each cable segment, the location of each segment and, therefore, the entire cable may be determined.
A system including a main neural network for performing one or more machine learning tasks on a network input to generate one or more network outputs. The main neural network includes a Mixture of Experts (MoE) subnetwork that includes a plurality of expert neural networks and a gating subsystem. The gating subsystem is configured to: apply a softmax function to a set of gating parameters having learned values to generate a respective softmax score for each of one or more of the plurality of expert neural networks; determine a respective weight for each of the one or more of the plurality of expert neural networks; select a proper subset of the plurality of expert neural networks; and combine the respective expert outputs generated by the one or more expert neural networks in the proper subset to generate one or more MoE outputs.
Providing at least one contextually relevant suggestion to one or more users of an ongoing message exchange thread between the users. The suggestion is provided for presentation to the user(s) via user interface output device(s) of computing device(s) of the user(s). The suggestion indicates a query that can be submitted to an automated assistant to cause the automated assistant to incorporate, into the message exchange thread, content that is responsive to the query. In some implementations, the suggestion is a selectable suggestion and content that is responsive to the query is incorporated into the message exchange thread in response to user interface input that is directed to the selectable suggestion. In some implementations, the suggestion is determined based on one or more messages that have already been communicated between users of the message exchange thread.
In various implementations, audio data that captures a spoken utterance of a first user and a spoken utterance of a second user is received. The audio data can be generated by microphone(s) of a transcription device. It can be determined, based on determining that the spoken utterance of the first user and the spoken utterance of the second user overlap for at least a threshold period of time, that the first user is a member of a first conversational cluster. The first conversational cluster includes at least one other participant and does not include the second user. A transcription can be generated based on performance of automatic speech recognition on the audio data and can include recognized text from the spoken utterance of the first user and can be annotated to indicate that such recognized text is part of the first conversational cluster.
G10L 17/02 - Preprocessing operations, e.g. segment selectionPattern representation or modelling, e.g. based on linear discriminant analysis [LDA] or principal componentsFeature selection or extraction
H04R 1/40 - Arrangements for obtaining desired frequency or directional characteristics for obtaining desired directional characteristic only by combining a number of identical transducers
In general, techniques are described that are directed to a device comprising an antenna, a radio, and processing circuitry. The radio may wirelessly communicate data, via the antenna. The processing circuitry may determine whether the device is contained within a case, and responsive to determining that the device is contained within the case, select a first proximity threshold as a current proximity threshold by which to determine whether a user is proximate to the device. The first proximity threshold may be different than a second, proximity threshold used responsive to determining that the device is not contained within the case. The processing circuitry may next cause, based on the selected proximity threshold, the radio to configure either a first transmission power or a second transmission power as a current transmission power. The radio may wirelessly communicate the data via the antenna using the current transmission power.
H04W 52/28 - TPC being performed according to specific parameters using user profile, e.g. mobile speed, priority or network state, e.g. standby, idle or non-transmission
H04W 52/36 - Transmission power control [TPC] using constraints in the total amount of available transmission power with a discrete range or set of values, e.g. step size, ramping or offsets
A user equipment (UE) receives, from a radio access network (RAN) in a serving cell, a delta configuration related to a target cell, for use in accessing the target cell subsequent to an activation command; receives, from the RAN, an activation command related to the delta configuration; and in response to the activation command the RAN, uses the delta configuration and at least a portion of a prior configuration to begin communicating on the target cell.
A device may receive a first indication that a head mounted device is being worn by a user, the head mounted device including a headset comprising a front frame portion connected to a first arm portion and a second arm portion, receiving a first measurement from a sensor indicating an arm portion to skin distance, receiving a second measurement from a inertial measurement unit indicating an arm portion orientation, executing a comfort prediction model using the first measurement and the second measurement as inputs to generate a predicted comfort score; and generating a second indication to display a headset fit instruction based on the predicted comfort score.
A bitstream that includes an encoded representation of a source image is obtained. A realism factor indicative of an amount of synthesized content in a reconstructed image of the source image is received. The realism factor and the encoded representation are input to a decoder to obtain the reconstructed image of the source image. The reconstructed image is store or displayed. The encoded representation can be a latent space representation of the source image and is created by an encoder. The realism factor can be derived from a range of values indicative of desired perceptual qualities in the reconstructed image.
A link layer architecture for D2D interconnections or communications. The architecture includes a self-contained D2D module that is capable of receiving data in multiple input bus protocols and transferring that data through a physical layer connection to another chiplet. The D2D module is a self-contained base block used to transfer data across dies. It includes a transmit (Tx) section/logic and a receive (Rx) section/logic. Each section/logic forms a building block that can be used to transfer data between dies or chiplets.
G06F 5/06 - Methods or arrangements for data conversion without changing the order or content of the data handled for changing the speed of data flow, i.e. speed regularising
G06F 1/06 - Clock generators producing several clock signals
G06F 13/20 - Handling requests for interconnection or transfer for access to input/output bus
A method may include obtaining a pretrained image encoder and a training sample comprising a training image and a training text string corresponding to the training image. The method may also include initializing a text encoder in an untrained state, determining, using the pretrained image encoder and based on the training image, a first latent representation of the training image, and determining, using the text encoder and based on the training text string, a second latent representation of the training text string. The method may further include determining a loss value based on the first latent representation and the second latent representation, updating, based on the loss value, one or more parameters of the text encoder while holding fixed parameters of the pretrained image encoder, and outputting the text encoder in a trained state.
Techniques and apparatuses are described for implementing dynamic thermal control for a system-on-chip. In example aspects, a thermal control system of a system-on-chip is a passive system capable of managing heat that is generated by a subsystem of the system-on-chip. The thermal control system dynamically triggers the subsystem to operate at different operation points to address thermal control requirements of the system-on-chip. In more detail, the thermal control system provides flexible and adaptive thermal control policy enforcement by adjusting a temperature threshold and/or feedback signal dynamics (e.g., a feedback time scale). The temperature threshold can be set based on a predicted operation of the subsystem or based on a current operation of the subsystem relative to one or more safety or performance limitations. Through dynamic thermal control, the thermal control system can enforce different throttling limits while the subsystem is operating under different workloads.
Implementations are directed to generating an edited synthetic image, that corresponds to a real image captured using a real camera, but that is generated based on, and reflects, edit(s) to an original natural language (NL) caption for the real image. For example, the original NL caption can be used in performing an inversion on the real image to generate a diffusion trajectory for the real image. Further, the diffusion trajectory can be used to optimize a sequence of unconditional embeddings, for the real image, that are not based on the NL caption for the real image. Yet further, the edited NL caption, the unconditional embeddings, and at least part of a noise vector (of the diffusion trajectory) can be processed, using a Large-scale language-image (LLI) model, to generate the edited synthetic image.
An electronic device is provided having an antenna that also functions as a display shield for a display of the device. The display shield can separate components for a display module from other electrical components of the electronic device. The display shield can be grounded to an enclosure of the device at least partially by one or more shield grounding clips, and configured to receive and/or transmit radio frequency waves.
A data center thermal control system includes a local cooler configured to cool a local coolant used for cooling electronic hardware, an outer heat exchanger configured to exchange heat from fluid to outside air, and a fluid circulation system configured to convey heat from the local cooler to the outer heat exchanger by circulating at least one fluid cooling medium, the fluid circulation system including a cold portion directed to the air cooler. The thermal control system also includes one or more processors and a non-transitory computer-readable medium storing instructions that, when executed by the one or more processors, would cause the one or more processors to govern the outer heat exchanger to cool fluid in the cold portion to a first target temperature during a hot season, and cool fluid in the cold portion to a lower target temperature during a cold season.
A central problem in training NeRF models is addressed, namely, optimization in the presence of distractors, such as transient or moving objects and photometric phenomena that are not persistent throughout the capture session. Example techniques formulate training as a form of iteratively re-weighted least squares, with a variant of trimmed LS, and an inductive bias on the smoothness of the outlier process.
Provided are computer-implemented systems and methods for responding to visual queries using both a visual search engine and a vision language model (VLM). In particular, aspects of the present disclosure can improve the performance of a VLM at generating a response to a visual queries by supplementing the visual query with one or more annotations generated by or using the visual search engine.
Encoding and decoding a block using power-of-two sizes is described. At a decoder, a current block to be decoded is identified from an encoded bitstream. The current block was partitioned by an encoder using a recursive partitioning scheme. The recursive partitioning scheme includes only a no split split-type, binary split-types, and at least two quad split-types that each result in at least two partitions of different dimensions. One or more sub-blocks are decoded to reconstruct the current block. The recursive partitioning scheme may be one of multiple recursive partitioning schemes. The recursive partitioning scheme, the available split-types of a recursive partitioning scheme, or both, may be identified or limited by the block size.
H04N 19/119 - Adaptive subdivision aspects e.g. subdivision of a picture into rectangular or non-rectangular coding blocks
H04N 19/176 - Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
H04N 19/192 - Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the adaptation method, adaptation tool or adaptation type used for the adaptive coding the adaptation method, adaptation tool or adaptation type being iterative or recursive
51.
FRAMEWORK FOR MANAGING MACHINE LEARNING PERCEPTION TASKS ON LIMITED HARDWARE RESOURCES
Techniques for managing machine learning (ML) perception tasks on limited hardware resources are disclosed herein. An example system includes one or more processors that execute instructions to: register, to a registry, for each of a plurality of ML task processors, an indication of a semantic output generated by the ML task processor; receive, by a task management module and from a client, a request for particular semantic output; identify, based on the registry, a ML task processor, wherein the indication of the semantic output for the ML task processor corresponds to the particular semantic output; responsive to determining the ML task processor is currently executing, refrain from loading the ML task processor; responsive to determining the ML task processor is not currently executing, execute the ML task processor; and establish a connection between the client and the ML task processor to allow the client to receive the semantic output.
Techniques include calibrating the user-facing sensors using a calibration target in world space. A calibration target in this case includes an LED panel that is mounted on a linear rail, and a jig for mounting a head-mounted display (HMD) having user-facing sensors such that the user-facing sensors face the LED panel. On the HMD, there are at least two user-facing cameras; an eye-tracking (ET) camera and a face-tracking (FT) camera. There are also a plurality of LEDs associated with an ET camera. In some implementations, the LEDs surround the ET camera. The cameras are calibrated simultaneously—that is, the camera intrinsics and extrinsics relative to a base, i.e., the LED panel in a base position, are determined in a single step after data has been collected. In another, separate step, the ET camera LEDs are calibrated—that is, their positions relative to the ET camera are determined.
Systems and methods for end-to-end pretraining of multimodal retrieval-augmented visual language models. In some examples, multimodal information may be encoded into key-value pairs and stored in a unified memory, which the model's retriever can access via multimodal query encodings in order to identify relevant information within multiple knowledge sources. The model may include an attentive fusion layer so that automatically-generated retrieval scores for multiple simultaneously-considered documents may b used in calculating attention scores, and gradients from the final task may be used to train the entire model (including the retriever) end-to-end and update the unified memory. In such cases, the retriever may thus be trained with the rest of the model without the need for ground-truth scores indicating which knowledge entries are most helpful in answering a given query, and the model's parameters may thus be focused on understanding queries and conducting reasoning rather than simply memorizing the training data.
A touch sensitive display assembly is provided. The touch sensitive display assembly includes a three-dimensional display cover defining an internal volume. The three-dimensional display cover includes a central portion and a peripheral portion. The peripheral portion extends around a periphery of the internal volume. The touch sensitive display assembly includes a display panel positioned within an internal volume defined by a three-dimensional display cover having a central portion and a peripheral portion extending around a periphery of the internal volume. The touch sensitive display further includes a first plurality of touch sensors operable to detect a touch input provided at the central portion of the three-dimensional display cover and a second plurality of touch sensors operable to detect a touch input provided at the peripheral portion of the three-dimensional display cover. and a second plurality of touch sensors.
Provided are computer-implemented systems and methods for responding to visual queries using both a visual search engine and a vision language model (VLM). In particular, aspects of the present disclosure can improve the performance of a VLM at generating a response to a visual queries by supplementing the visual query with one or more annotations generated by or using the visual search engine.
G06F 16/583 - Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
09 - Scientific and electric apparatus and instruments
38 - Telecommunications services
42 - Scientific, technological and industrial services, research and design
Goods & Services
Downloadable software for publishing and sharing digital
media and information via global computer and communication
network; downloadable instant messaging software;
downloadable communications software for electronically
exchanging data, video and graphics accessible via computer,
mobile, wireless, and telecommunication networks;
downloadable computer software for processing images,
graphics, video, and text; downloadable video and audio
conferencing software. Telecommunications services, namely, electronic transmission
of data and digital messaging via global computer and
communication networks; instant messaging services; video
and audio conferencing services conducted via the web, and
mobile devices; communications by computer terminals; mobile
telephone communication services. Providing non-downloadable software for publishing and
sharing digital media and information via global computer
and communication network; providing non-downloadable
instant messaging software; providing non-downloadable
communications software for electronically exchanging data,
video and graphics accessible via computer, mobile,
wireless, and telecommunication networks; providing
non-downloadable computer software for processing images,
graphics, video, and text; providing non-downloadable video
and audio conferencing software.
42 - Scientific, technological and industrial services, research and design
Goods & Services
Providing on-line non-downloadable computer software to
create and edit surveys, quizzes, and forms to collect and
analyze data; providing on-line non-downloadable computer
software for tracking documents over computer networks,
intranets and the Internet; providing on-line
non-downloadable software for document collaboration and
revision tracking; providing on-line non-downloadable
software for granting and controlling access to documents;
providing on-line non-downloadable software featuring online
storage of documents and databases.
60.
SYSTEMS AND METHODS FOR PROVIDING REAL-TIME FEEDBACK VIA GRAPHICAL USER INTERFACE ELEMENTS
Systems and methods for providing real-time feedback via a graphical user interface are provided. Example techniques may include providing, by a user device, a user interface including a graphical display associated with a transaction, wherein the graphical display includes: (i) an interactive slider element configured to visually represent completing the transaction based on a user action including an interaction with the interactive slider element and (ii) one or more additional elements configured to provide additional user feedback based on the user action; detecting, by the user device, an initiation of the user action with respect to the slider element and the one or more additional elements of the graphical display; and providing, by the user device, prior to completion of the user action, real-time feedback via the slider element and the one or more additional elements in response to the initiation of the user action of the graphical display.
G06F 3/01 - Input arrangements or combined input and output arrangements for interaction between user and computer
G06F 3/0481 - Interaction techniques based on graphical user interfaces [GUI] based on specific properties of the displayed interaction object or a metaphor-based environment, e.g. interaction with desktop elements like windows or icons, or assisted by a cursor's changing behaviour or appearance
G06Q 20/32 - Payment architectures, schemes or protocols characterised by the use of specific devices using wireless devices
There is disclosed a method for federated learning implemented by a client device. The method comprises obtaining a model input and processing, by a neural network comprising one or more neural network layers, the model input to generate a model output. The method further comprises caching the activations of one or more selected layers of the neural network generated from processing the model input by the neural network and generating a label based upon a user interaction with the model output. The method further comprises transmitting, to a server, training data based upon the cached activations and corresponding labels for generating one or more steering vectors for the one or more selected layers.
A computer-implemented method includes detecting, by one or more processors of a first computing system, a machine-readable encoding generated by a second computing system. The machine-readable encoding includes an embedded prompt configured for processing by one or more machine-learned models. Further, the method includes providing, by one or more processors of the first computing system, at least the prompt as input to a generative model stored by the first computing system. Moreover, the method includes obtaining, by the one or more processors of the first computing system, scripted language code as output from the generative model in response to the prompt. In addition, the method includes generating, by the one or more processors of the first computing system, a lightweight application based on the scripted language code.
Methods, systems, and apparatus for a computing device comprising; a plurality of processing cores; and a reservation station comprising circuitry configured to coordinate the selection of instructions for out-of-order execution on the plurality of processing cores, wherein the reservation station comprises a waiting buffer and a plurality of clusters, wherein upon the reservation station predicting that a load instruction will result in a cache miss, the reservation station is configured to execute the load instruction using a cluster of the plurality of clusters and to store one or more dependent instructions of the load instruction in the waiting buffer, and wherein upon the load instruction completing execution, the reservation station is configured to obtain the dependent instructions from the waiting buffer and execute the dependent instructions using the plurality of clusters.
The present disclosure is directed to generating lightweight (e.g., distilled) representations of training data sets for on-device personalized learning. Distilled training examples are used as a regularizer for personalized learning. Personalized learning involves locally fine-tuning a model with user examples. The embodiments deploy of a machine learning (ML) model (e.g., a generative model) that procedurally generates training samples that closely approximates the data (probability distribution) of the training set. More specifically, the model generates a “distilled” personalized training data set to be employed locally for on-device personalized learning of a generalized trained model. Because the ML model (deployed to the model for personalized training of a target model) generates a distilled training data set, the ML model may be referred to as a training set distillation (TSD) model.
Example implementations provide a computer-implemented method for training a machine-learned model, the method comprising: processing, using a layer of the machine-learned model, positive input data in a first forward pass; updating one or more weights of the layer to adjust, in a first direction, a goodness metric of the layer for the first forward pass; processing, using the layer, negative input data in a second forward pass; and updating the one or more weights to adjust, in a second direction, the goodness metric of the layer for the second forward pass.
In general, the subject matter described in this disclosure can be embodied in methods, systems, and program products for detecting a mobile device pointing gesture toward another device. A mobile device determines that the mobile device physically moved in a manner that satisfies criterion for a pointing gesture. The mobile device or another system device determines that the pointing is oriented toward a particular device, including by receiving a signal using a first antenna and a second antenna, and by determining an indication of a signal difference between receipt of the signal by the first antenna and the second antenna.
G08C 17/02 - Arrangements for transmitting signals characterised by the use of a wireless electrical link using a radio link
G01S 3/48 - Systems for determining direction or deviation from predetermined direction using antennas spaced apart and measuring phase or time difference between signals therefrom, i.e. path-difference systems the waves arriving at the antennas being continuous or intermittent and the phase difference of signals derived therefrom being measured
G06F 3/01 - Input arrangements or combined input and output arrangements for interaction between user and computer
G06F 3/0346 - Pointing devices displaced or positioned by the userAccessories therefor with detection of the device orientation or free movement in a 3D space, e.g. 3D mice, 6-DOF [six degrees of freedom] pointers using gyroscopes, accelerometers or tilt-sensors
After authenticating a user, a computing device detects, by a fingerprint sensor, a touch input provided within a. predetermined period of time of the computing device authenticating the user. While detecting the touch input, the fingerprint sensor captures fingerprint information. The computing device determines whether the fingerprint information corresponds to any enrolled fingerprint from a set of enrolled fingerprints stored at the computing device. Responsive to determining that the fingerprint information does not correspond to any enrolled fingerprint from the set of enrolled fingerprints, the computing device stores the fingerprint information as an enrolled non-fingerprint.
Systems and methods for data compression are disclosed. A conventional codec that uses multiple prediction schemes encodes a predictor indicator for each data point, adding data overhead to the output stream. The disclosed technology reduces this overhead by conditionally omitting the predictor indicator. An encoder generates multiple predicted values for a data point using different prediction schemes. The encoder compares these values against a criterion, such as whether two or more values are equivalent. If the criterion is met, the predictor indicator is omitted from the output stream. A decoder, configured with the same deterministic logic, applies the same comparison to determine if a predictor indicator is present in the output stream and thereby identifies which predictor was used. Conditionally omitting the predictor indicator creates a more compact output stream and achieves an improved compression rate for the same reconstructed data.
Implementations of the present disclosure include actions of receiving image data of an image capturing a scene, receiving data describing one or more entities determined from the scene, the one or more entities being determined from the scene, determining one or more actions based on the one or more entities, each action being provided at least partly based on search results from searching the one or more entities, and providing instructions to display an action interface comprising one or more action elements, each action element being to induce execution of a respective action, the action interface being displayed in a viewfinder.
G09G 5/36 - Control arrangements or circuits for visual indicators common to cathode-ray tube indicators and other visual indicators characterised by the display of individual graphic patterns using a bit-mapped memory
G06F 3/04817 - Interaction techniques based on graphical user interfaces [GUI] based on specific properties of the displayed interaction object or a metaphor-based environment, e.g. interaction with desktop elements like windows or icons, or assisted by a cursor's changing behaviour or appearance using icons
G06F 3/04842 - Selection of displayed objects or displayed text elements
Interface and mode selection for digital action execution is provided. For example, a system loads a script library embedded in an electronic resource. The system determines a historic level of engagement between a client computing device and one or more digital assistants. The system selects, based on a first property of the client computing device and the historic level of engagement, a type of digital interface. The system generates, based on the type of digital interface, a digital interface with the call-to-action. The system determines, responsive to an instruction to execute the call-to-action, a mode of execution. The system selects a digital assistant and a second client device to execute the call-to-action. The system transmits the call-to-action to the second client device for execution.
Implementations relate to restricting access of an application to audio data content captured subsequent to rendering content to the user at the request of the application. An application can generate content that is to be rendered to a user with an additional request to receive audio data content from audio data captured immediately after rendering the content. The content can be processed, using a trained machine learning model that generates, as output, an indication of likelihood that providing audio data content after rendering the content from the application was improper. In instances the application improperly requested audio data content, the application can be restricted from being provided the audio data content and/or subsequent audio data content.
Systems and methods of selecting digital platforms for execution of voice-based commands are provided. The system receives an application that performs an action associated with a service via digital platforms. The system debugs the application to validate parameters of the action on at least two platforms of the digital platforms. The system receives data packets comprising an input audio signal detected by a sensor of a client device, and parses the input audio signal to identify the action and the service. The system selects a first platform from the digital platforms to perform the action. The system initiates, responsive to selection of the first platform, an interactive data exchange to populate parameters of an action data structure corresponding to the action. The system executes the action via the selected platform using the action data structure.
A method of performing a reshape operation specified in a reshape layer of a neural network model is described. The reshape operation reshapes an input tensor with an input tensor shape to an output tensor with an output tensor shape. The tensor data that has to be reshaped is directly routed between tile memories of the hardware accelerator in an efficient manner. This advantageously optimizes usage of memory space and allows any number and type of neural network models to be run on the hardware accelerator.
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing computational graphs on distributed computing devices. One of the methods includes receiving a request to execute a processing pipeline (i) first operations that transform raw inputs into pre-processed inputs and (ii) second operations that operate on the pre-processed inputs; and in response: assigning the first operations to two or more of a plurality of computing devices, assigning the second operations to one or more hardware accelerators of a plurality of hardware accelerators, wherein each hardware accelerator is interconnected with the plurality of computing devices, and configured to (i) receive inputs from respective queues of the two or more computing devices assigned the first operations and (ii) perform the second operations on the received pre-processed inputs, and executing, in parallel, the processing pipeline on the two or more computing devices and the one or more hardware accelerators.
Implementations relate to low power consumption for hazard detection by mobile devices. In some implementations, a computer-implemented method includes obtaining sensor data from sensors of a device, including global positioning sensor data indicating geographical locations of the device. Map data is obtained that describes features of a geographical area in which the device is located. Based on the sensor data, map data, and the geographic locations, it is determined whether one or more potential hazards are within a threshold distance of the device. If so, a rear camera of the device is activated, images are captured using the rear camera, and it is determined whether hazard objects are positioned within a particular distance of the device based on the captured images. An alert is output by the processor in response to determining that hazard object(s) are within the particular distance of the device.
The present disclosure provides systems and methods to measure quantum gate fidelity through swap spectroscopy. In particular, aspects of the present disclosure are directed to the derivation and use of a physical model that models non-Markovian quantum dynamics of interactions between one or more qubits of a quantum gate and one or more two-level-system (TLS) defects during operation of the quantum gate.
G06N 10/00 - Quantum computing, i.e. information processing based on quantum-mechanical phenomena
G01N 21/31 - Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
G06N 7/01 - Probabilistic graphical models, e.g. probabilistic networks
Techniques and devices for low-latency connection establishment to devices that sleep (reduce power consumption including powering-down radio hardware) for periods of time. A first node (302) transmits (306), to a second node (304), one or more Wake frames, and in response to the transmission of the one or more Wake frames, the first node receives a Link Request message (308) from the second node. Based on receiving the Link Request message, the first node transmits a Link Accept and Request message (312) to the second node. In response to transmitting the Link Accept and Request message, the first node receives a Link Accept message (316) from the second node that is effective to establish the wireless peer-to-peer communication link between the first node and the second node.
Aspects of the disclosed technology include machine-learning systems and methods for enhancing speech signals. A machine-learned speech enhancement model can be trained using audio data and synthetic accelerometer data that is generated using a synthetic accelerometer data generation model. The accelerometer data can include synthetic noise data that can be used to train the speech enhancement model to learn a transfer function that models environmental noise to an accelerometer of a wearable device. By simulating data from multiple sensors, the disclosed technology provides a cost-effective solution for training robust large-scale models for speech enhancement.
G10L 25/30 - Speech or voice analysis techniques not restricted to a single one of groups characterised by the analysis technique using neural networks
Methods, apparatuses, and non-transitory computer-readable media for wireless communication. One method includes detecting that multiple single-address network connections are connected between the UE and a same first third generation partnership project (3GPP) data network access point, and in response to detecting that the multiple single-address network connections are connected between the UE and the same first 3GPP data network access point, establishing a multi-address network connection between the UE and the same first 3GPP data network access point, to merge the multiple single-address network connections between the UE and the first 3GPP data network access point to the multi-address network connection between the UE and the first 3GPP data network access point.
H04W 76/15 - Setup of multiple wireless link connections
H04W 8/26 - Network addressing or numbering for mobility support
H04W 76/16 - Setup of multiple wireless link connections involving different core network technologies, e.g. a packet-switched [PS] bearer in combination with a circuit-switched [CS] bearer
H04W 88/06 - Terminal devices adapted for operation in multiple networks, e.g. multi-mode terminals
80.
SHORTENING A SETUP TIME OF A DATA CONNECTION SETUP PROCEDURE ON A USER EQUIPMENT
Methods, apparatuses, and non-transitory computer-readable media for wireless communication. One method includes obtaining a first internet protocol (IP) prefix address of a first data connection with an access point of a wireless network, and in response to the first data connection being disconnected, establishing a second data connection with the access point of the wireless network based on the first IP prefix address, performing an internet control message protocol for IP version 6 (ICMPv6) router solicitation and router advertisement (RS/RA) procedure to obtain a second IP prefix address, and determining whether an IP change procedure is to be performed for the second data connection based on a comparison between the first IP prefix address and the second IP prefix address.
Techniques and devices for dynamic router selection for robust wireless mesh communications are described for a joining device that performs a discovery procedure to join the wireless mesh network, and based on discovery responses from the discovery procedure, measures a bidirectional link margin for each discovery response received from a router in the wireless mesh network. The joining device compares measured bidirectional link margins to a minimum-bidirectional-link-margin-threshold. If no measured bidirectional link margins for routers exceed the minimum-bidirectional-link-margin-threshold, the joining device compares a link-quality to neighboring routers, received from one or more router-eligible end devices during the discovery procedure, to a minimum-bidirectional-link-margin-threshold-for-router-to-router-communication. The joining device selects a router or a router-eligible end device to use to join the wireless mesh network based on the evaluation of the link qualities.
Techniques and apparatuses are described that perform a determination of craniofacial gestures of a wearer of an on-head wearable. In an example aspect, an intelligent personal assistant device 110 can proactively engage with users based on their craniofacial gestures detected through an on-head wearable device 108, such as earbuds. Using various sensors, the wearable device monitors combinations of head position and facial expressions to anticipate when the wearer might need assistance. The wearable device's sensors monitor head and facial positions and movements, and when they detect a gesture that matches one associated with assistant engagement, they notify the intelligent personal assistant 110 to interact with the wearer. This system allows the assistant to initiate contact at potentially helpful moments rather than waiting for direct commands. Additionally, various craniofacial gestures can trigger various other actions through the device.
Methods, systems, and apparatus for predicting errors that occurred during an implementation of an error-corrected transversal CNOT gate between a first logical qubit and a second logical qubit. In one aspect, a method includes decoding first measurement data obtained from measure qubits included in the first logical qubit to detect occurrences of a first error type in the first logical qubit; decoding second measurement data from measure qubits included in the second logical qubit to detect occurrences of a second error type in the second logical qubit; copying detected occurrences of the second error type to the first measurement data to generate updated first measurement data, comprising removing detection events in the first measurement data that correspond to the detected occurrences of the second error type; and decoding the updated first measurement data to detect occurrences of the second error type in the first logical qubit.
84.
FOLDABLE ELECTRONIC DEVICE WITH DETUNE ELEMENTS INTEGRATED IN AN NFC ANTENNA LOOP FOR CONTROLLED ANTENNA DECOUPLING WHEN CLOSED
A foldable electronic device (102) comprises a housing (104) having a first housing section (112) and a second housing section (114) connected by a movement structure (116), such as a hinge, allowing the housing to rotate between an open orientation, in which the facing surfaces (120, 124) of the housing sections are substantially coplanar, and a closed orientation, in which the surfaces (120, 124) are substantially parallel and positioned on the same side of the movement structure (116). One or more display modules (106-1, 106-2, 106-3) are arranged on one or both housing sections (112, 114). A plurality of antennas (306) are located along a first bezel (304) of the first housing section (112). A plurality of detune elements (134) are disposed along a second bezel (308) of the second housing section (114) and are configured to influence or detune the resonance of the antennas (306) when the device (102) is folded. The detune elements (134) are further incorporated into a near-field communication (NFC) antenna loop (136) positioned on the second bezel (308), enabling NFC transmission and reception. By electrically coupling the detune elements (134) with the antennas (306) through the housing geometry, the foldable device maintains stable antenna impedance, reduces electromagnetic coupling between the housing sections, and preserves reliable NFC communication efficiency in both the open and closed states.
A method (500) for providing personalized responses to queries using a personalized large language model (LLM) includes receiving a query (202) from a user specifying a task for an assistant LLM (240) to perform, the query captured by an assistant-enabled device associated with the user, and processing the query to identify, from a datastore (230) of a plurality of embedding chunks (342) each previously stored in the datastore by the assistant LLM, a particular embedding chunk that is relevant to the query. The method also includes generating an on-the-fly prompt (222) by stitching the particular embedding chunk and the query together, and processing, by the assistant LLM, the on-the-fly prompt to generate a personalized response (242) to the query.
An example method includes analyzing, by a camera assistant machine learning (ML model, a scene displayed in a preview of an image capturing device. The method includes identifying, by the camera assistant ML model and based on the analyzing of the scene, image capture characteristics of an environment of the scene. The method includes receiving, from a user, a selection of a virtual photographer persona comprising style attributes of a real photographer. The method includes generating, by the camera assistant ML model and based on the image capture characteristics and the selected virtual photographer persona, a collection of image capture guidelines to guide the user to capture an image or a video of the scene. Upon capture, the image or the video corresponds to the style attributes of the real photographer. The method includes providing, during the image capturing process, the generated collection of image capture guidelines.
A method of wireless communication at a central unit (CU) of a network entity is provided. According to the method, the CU transmits, to a distributed unit (DU) of the network entity, a first CU-to-DU message including at least one of: a lower layer triggered mobility (LTM) indication, a cell identifier (ID) of a cell, or cell group configuration information. The CU receives, from the DU in response to the first CU-to-DU message, a first DU-to-CU message including a LTM master cell group (MCG) configuration. The CU transmits, to the DU, a radio resource control (RRC) reconfiguration message including a LTM candidate configuration based on the LTM MCG configuration. A method at a DU is also provided. The methods can be implemented on an apparatus having a memory, a transceiver, and a processor.
09 - Scientific and electric apparatus and instruments
42 - Scientific, technological and industrial services, research and design
Goods & Services
Downloadable cloud computing software for use in creating,
storing, and editing text documents, spreadsheets, tables,
slides, drawings, images, audio files, video files, scanned
documents, and websites. Cloud computing featuring software for use in creating,
storing, and editing text documents, spreadsheets, tables,
slides, drawings, images, audio files, video files, scanned
documents, and websites; application service provider (ASP),
namely, hosting computer software applications of others;
application service provider (ASP) featuring software for
use in creating, storing, and editing text documents,
spreadsheets, tables, slides, drawings, images, audio files,
video files, scanned documents, and websites; computer
services, namely, integration of private and public cloud
computing environments; computer services, namely, cloud
hosting provider services; providing virtual computer
systems and virtual computer environments through cloud
computing; technical support services, namely,
troubleshooting of computer software problems in the field
of cloud computing; application service provider, namely,
hosting and maintenance of collaborative websites created by
others using non-downloadable software.
09 - Scientific and electric apparatus and instruments
38 - Telecommunications services
42 - Scientific, technological and industrial services, research and design
Goods & Services
Downloadable software for publishing and sharing digital
media and information via global computer and communication
network; downloadable instant messaging software;
downloadable communications software for electronically
exchanging voice, data, video and graphics accessible via
computer, mobile, wireless, and telecommunication networks;
downloadable computer software for processing images,
graphics, audio, video, and text; downloadable video and
audio conferencing software; Computer hardware, computers,
video monitors, audio speakers, microphones, speaker
microphones, video cameras. Telecommunications services, namely, electronic transmission
of data and digital messaging via global computer and
communication networks; instant messaging services; video
and audio conferencing services conducted via the web,
telephone, and mobile devices; communications by computer
terminals; local and long distance telephone services;
mobile telephone communication services. Providing temporary use of on-line non-downloadable software
for publishing and sharing digital media and information via
global computer and communication networks; providing
temporary use of on-line non-downloadable software
development tools; providing a web hosting platform for
others for organizing and conducting meetings, social events
and interactive text, audio, and video discussions;
providing an on-line network environment that features
technology that enables users to share data; computer
software consulting; application service provider (ASP)
services featuring computer software for transmission of
text, data, images, audio, and video by wireless
communication networks and the Internet; application service
provider (ASP) services featuring computer software for
electronic messaging and wireless digital messaging.
42 - Scientific, technological and industrial services, research and design
Goods & Services
Providing on-line non-downloadable software for use in
collaboration, editing, and creating videos; providing
on-line non-downloadable software for granting and
controlling access to videos; providing on-line
non-downloadable software featuring online storage of videos
and databases.
09 - Scientific and electric apparatus and instruments
42 - Scientific, technological and industrial services, research and design
Goods & Services
Downloadable software for capturing, storing, accessing,
displaying, sharing and managing digital files, including
text, graphics, and images; downloadable software for
aggregating, organizing, and interacting with digital
content, including text, graphics, and images. Providing the temporary use of online non-downloadable
software for capturing, storing, accessing, displaying,
sharing and managing digital files, including text,
graphics, and images; providing the temporary use of online
non-downloadable software for aggregating, organizing and
interacting with digital content, including text, graphics,
and images.
09 - Scientific and electric apparatus and instruments
38 - Telecommunications services
42 - Scientific, technological and industrial services, research and design
Goods & Services
Downloadable computer software for providing electronic mail
services. Electronic transmission of e-mail and messages. Providing non-downloadable computer software for providing
electronic mail services.
This document describes systems and techniques directed at cache-optimized warp engines. In aspects, a computing device having a first cache memory, a second cache memory, a system memory, and a cache-optimized warp engine is configured to receive a warped input image comprising a plurality of pixel information. Following a coordinate sequence, the cache-optimized warp engine scans the plurality of pixel information and loads a first portion of the pixel information into the first cache memory and a second portion of the pixel information into the second cache memory. Based on the first and second portions of the plurality of pixel information, the cache-optimized warp engine determines first and second portions, respectively, of a plurality of pixel information of a corrected output image. The cache-optimized warp engine stores the first and second portions of the pixel information of the corrected output image as the output image in the system memory.
Example embodiments of the present disclosure provide for an example method for prompt element generation for use as input into generative models. The example method includes obtaining user input data including initial prompt data. The example method includes selecting one or more suggested prompt elements based at least in part on the initial prompt data. The example method includes transmitting the one or more suggested prompt elements to be presented for display as selectable user interface elements via a user interface. The example method includes obtaining second user input data including data indicative of a selection of the one or more suggested prompt elements. The example method includes responsive to obtaining the second user input data, providing the second user input to an image generation model to generate an output image including a visual representation associated with the one or more suggested prompt elements.
Methods, systems, and apparatus, including medium-encoded computer program products for selecting and displaying content in privacy preserving manners are described. A digital component request that includes contextual data related to an environment in which the digital component will be displayed can be received from a client device by a first content platform. Based on the contextual data, the user can be assigned to user attribute buckets, which can be associated with at least one type of user attribute. Based on the contextual data and each user attribute bucket to which the user is assigned, candidate digital components can be selected for distribution to the client device. The application can be provided response data that initiates an update to aggregated user attribute data for the user and enables the application to select a digital component to display to the user based on the aggregated data.
A waveguide combiner includes an incoupler, an outcoupler, and a plurality of exit pupil expanders. The incoupler is configured to split incoming display light into multiple light beams. Each exit pupil expander of the plurality of exit pupil expanders includes one or more facets configured to receive a corresponding light beam of the multiple light beams and to output light towards the outcoupler based on the received corresponding light beam.
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating embeddings of spoken utterances. One of the methods includes obtaining audio data representing a spoken utterance; processing the audio data using an encoder neural network to generate an embedding of the spoken utterance; and processing the embedding of the spoken utterance using a prediction neural network to generate a prediction about the spoken utterance, the processing comprising: maintaining respective embeddings for a plurality of preceding spoken utterances; determining one or more embeddings of respective preceding spoken utterances that are relevant to generating the prediction about the spoken utterance; and processing (i) the embedding of the spoken utterance and (ii) the respective embeddings of the one or more determined preceding spoken utterances to generate the prediction about the spoken utterance.
A computing device for generating a travel itinerary includes one or more memories configured to store instructions and one or more processors configured to execute the instructions to perform operations. The operations include receiving an input from a user relating to a query associated with a travel plan associated with a geographic area; based on the input, implementing one or more machine-learned models to generate a travel itinerary satisfying the query, wherein the one or more machine-learned models are trained to generate travel itineraries based on a plurality of images of maps associated with a plurality of geographic areas; and providing, as an output, the travel itinerary.
A method including sensing, by a wearable device, device context as a contextual cue, selecting, by the wearable device, a trained input model based on the contextual cue, generating, by the wearable device, a feature representation based on a wake-up cue using the input model, and predicting, by the wearable device, a wake-up trigger based on the feature representation.
G06F 3/02 - Input arrangements using manually operated switches, e.g. using keyboards or dials
G06N 3/0442 - Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
Aspects of the disclosure may involve training a sound event detection model (100, 300) to identify whether a second sound recording includes a sound of a reference clip. For instance, a sound event reference example including a first sound recording and a label indicative of whether the first sound recording includes a sound may be received. A breadth parameter, the breadth parameter being indicative of whether the sound event detection model detects a specific sound event, a class of sounds, or both may be received. The sound event reference example may be augmented to generate a test clip. The sound event detection model may be trained using the sound event reference example, the test clip, the breadth parameter, and the label. The sound event detection model includes a neural network (104, 304a, 304b) including reference encoder (106, 306a, 306b) and a sound event detector (108, 308a, 308b). The training may involve simultaneously training the neural networks.
