Techniques for altering audio being output by a voice-controlled device, or another device, to enable more accurate automatic speech recognition (ASR) by the voice-controlled device. For instance, a voice-controlled device may output audio within an environment using a speaker of the device. While outputting the audio, a microphone of the device may capture sound within the environment and may generate an audio signal based on the captured sound. The device may then analyze the audio signal to identify speech of a user within the signal, with the speech indicating that the user is going to provide a subsequent command to the device. Thereafter, the device may alter the output of the audio (e.g., attenuate the audio, pause the audio, switch from stereo to mono, etc.) to facilitate speech recognition of the user's subsequent command.
Network pathways are identified to transfer packets between a pair of regional virtual traffic hubs of a provider network. At a first hub of the pair, a first action is performed, resulting in a transmission of a packet received from a first isolated network to the second hub along a pathway selected using dynamic routing parameters. At the second hub, a second action is performed, resulting in the transmission of the packet to a destination within a second isolated network.
A network device can include processing circuitry to provide support for packet processing functions. The packet processing circuitry can perform egress operations such as encapsulating and segmenting egress data traffic from a virtual machine. The packet processing circuitry can also perform ingress operations such as coalescing and decapsulating ingress data traffic from a network.
H04L 41/082 - Configuration setting characterised by the conditions triggering a change of settings the condition being updates or upgrades of network functionality
Described are systems and methods for providing an encrypted LIDAR system, which may be used in autonomous vehicles. Embodiments of the present disclosure can provide LIDAR systems and methods employing a cryptographically secure deterministic random bit generator (DRBG) to generate a secure pseudorandom key sequence. The secure pseudorandom key sequence can be encoded into light pulses emitted by the exemplary systems and methods. Subsequently, upon receipt of reflected light, the exemplary systems and methods can authenticate the received reflected light to confirm the key sequence encoded in the received reflected light. If the key sequence encoded in the received reflected light matches the key sequence encoded into the emitted light pulse, parameters associated with the emitted and reflected light can be used to determine a distance to the object that may have reflected the emitted light pulse. In the event the received reflected light does not include the encoded key sequence the received light can be “discarded” by the exemplary systems and methods.
Disclosed are various embodiments for creating personalized insoles designed to provide physical comfort to the individual wearing shoes containing the personalized insoles. Three-dimensional (3D) foot scans are performed on the individual's feet to better understand the contours and geometry of each foot of the individual. The foot scans can be of each foot of the individual in varying positions (e.g., resting position with toes engaged with scanner top, toes-raised position with toes lifted from scanner top, metatarsal doming position, neutral calcaneal stance position, etc.). The 3D foot scans are used to generate personalized insole data that can be transmitted to a manufacturing device for manufacturing a personalized insole that is accurately designed to conform with the individual's foot and provide support in aligned load bearing positions.
Systems and methods for data transmission protocols are disclosed. For example, a receiving device is selected to send retransmission requests when data packet loss occurs, and then a unicast transmission of data packets may be sent to that selected receiving device. Other receiving devices may be configured to receive the unicast transmission as sent to the selected receiving device and to process the transmission as a multicast transmission. Alternatively, a multicast transmission may be sent to all of the receiving devices, and the selected receiving device may be configured to process the multicast transmission as a unicast transmission for the purpose of sending retransmission requests indicating packet loss. Backup packets may be generated and sent to all of the receiving devices to improve content output quality.
Methods of video streaming are generally described. In some examples, a camera device periodically captures an image, communicates encoded frame data representing that image to a server, and decodes and stores the previously encoded frame data as a background picture. The server receives the encoded frame data, decodes it, and stores the decoded frame in a buffer for future use. Subsequently, upon initiation of a streaming session, the camera device captures another image and encodes a predicted key frame based on differences between the captured image and the background picture. The camera device sends the predicted key frame to the server, which receives it and reconstructs a facsimile of the captured image utilizing the previously decoded frame stored in the buffer. Methods of acknowledging successfully decoded frames for use in selecting background pictures is also described.
H04N 19/593 - Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving spatial prediction techniques
H04N 19/139 - Analysis of motion vectors, e.g. their magnitude, direction, variance or reliability
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 19/625 - Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding using discrete cosine transform [DCT]
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
A system obtains a graph representing a set of resources of a distributed system. At least one node of the graph represents a resource-metric pair. The system further obtains time series data that indicates anomalies from the system. Then, the system determines a root cause anomaly that caused other anomalies based at least in part on the graph and the time series data.
H04L 43/067 - Generation of reports using time frame reporting
H04L 41/0631 - Management of faults, events, alarms or notifications using root cause analysisManagement of faults, events, alarms or notifications using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis
H04L 43/0817 - Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability by checking functioning
Techniques for allocating resources in an item handling environment are described. In an example, a computer system determines that an item is to be transferred at an item handling station to or from a stowage unit. The computer system determines data associated with transferring the item. The data indicates at least one of: an item type, a sequence of transferring items to or from the stowage unit, capability at the item handling station associated with handling the item type, an ergonomic parameter associated with the handling, a configuration of the stowage unit, an allocation of one or more resources to handle items, or a schedule of the allocation. The computer system generates, by using the data as input to an artificial intelligence model, a set of instructions indicating an allocation of a resource to the item handling station for the transferring of the item and a timing of the allocation.
Techniques for adaptive sleep virtual machine management are described. A service of a provider network receives a parameter indicating that a first virtual machine can be slept and determines to sleep the first virtual machine based at least in part on the parameter. A state of the first virtual machine is captured, and the first virtual machine is terminated. The service determines to resume the captured state of the first virtual machine based at least in part on an indication. A second virtual machine is launched using the captured state of the first virtual machine and resumes execution of the captured state of the first virtual machine. A proxy server sends traffic to the first virtual machine before the termination of the first virtual machine and to the second virtual machine after the resumption of execution of the captured state of the first virtual machine by the second virtual machine.
Techniques for in-flight scaling of machine learning training jobs are described. A request to execute a machine learning (ML) training job is received within a provider network, and the ML training job is executed using a first one or more compute instances. Upon a determination that a performance characteristic of the ML training job satisfies a scaling condition, a second one or more compute instances are added to the ML training job while the first one or more compute instances continue to execute portions of the ML training job.
Systems and methods are disclosed to perform gradient exchange among processing nodes configured as a hyper-rectangle network of N-dimensions. Each processing node can operate as a collective parameter server node capable to perform collective compute operations. For each dimension in a sequence of dimensions, all processing nodes on a same edge can perform a scatter-reduce operation using respective collective parameter serving engines. The amount of data reduced in each dimension is an inverse of a number of processing nodes in that dimension. After the scatter-reduce operation is performed for all the dimensions, all processing nodes on the same edge can perform an all-gather operation for each dimension in a reverse sequence of dimensions.
Devices, systems, and methods are provided for smart shopping based on recognition of objects presented in video. A method may include identifying, by a first device, using a machine learning model using a computer vision technique, objects represented in video content; determining that an object of the objects is available for purchase using an online retail system; causing concurrent presentation of the video content and a first indication that the object is available for purchase using the online retail system; receiving, from a second device, a second indication of a user selection of the first indication, wherein the user selection is indicative of a request to present additional information associated with the object; generating, based on the request, presentation data including the additional information and an option to purchase the object using the online retail system; and causing, based on the request, presentation of the presentation data.
Technologies directed to respiration monitoring based on noise channel state information (CSI) data are described. A method includes receiving CSI data representing channel properties of a wireless channel used by a first wireless device and a second wireless device located in a geographical region. The method generates a set of CSI samples by sampling the CSI data and removes one or more outlier CSI samples using a sparse outlier process, and removes a cluster of outlier samples using a cluster outlier process. The method determines Fast Fourier Transform (FFT) data for each channel subcarrier index and a signal-to-noise ratio (SNR) value for each channel subcarrier index, and identifies a first number of channel subcarrier indexes having highest SNR values to obtain a subset of FFT data, which represents a breathing spectrum. The method determines a respiratory rate of a user in the geographical region from the subset of the FFT data.
H04W 4/38 - Services specially adapted for particular environments, situations or purposes for collecting sensor information
H04B 7/06 - Diversity systemsMulti-antenna systems, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
H04B 17/336 - Signal-to-interference ratio [SIR] or carrier-to-interference ratio [CIR]
H04W 4/021 - Services related to particular areas, e.g. point of interest [POI] services, venue services or geofences
H04W 52/52 - Transmission power control [TPC] using AGC [Automatic Gain Control] circuits or amplifiers
A system configured to determine an estimated angle of arrival in reverberant environments. When a first device detects a calibration tone generated by a second device, the first device may generate multichannel audio representing the calibration tone and process the multichannel audio using a combination of detection filtering and subspace processing to determine a relative direction of the second device. For example, the first device may perform matched filtering to isolate a direct-path peak for the calibration tone, and then may sweep through all potential azimuth directions to identify an azimuth value corresponding to the direct-path peak. In some examples, the first device identifies a steering vector associated with a particular direction (e.g., signal subspace) that minimizes components in all other directions (e.g., noise subspace). The device may determine this steering vector independently for each frequency band and calculate the estimated angle of arrival by averaging results across frequency bands.
G01S 3/80 - Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received using ultrasonic, sonic, or infrasonic waves
G01S 3/801 - Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received using ultrasonic, sonic, or infrasonic waves Details
An input device determines presence of an actual user, instead of an artifact, by using multi-wavelength reflectance spectroscopy. Light sources are operated to illuminate an object with different colors of light at different times. A detector determines, at those different times, intensity data indicative of intensity light of these different colors as reflected from the object. The intensity data is processed to determine whether the object is part of a user or is an artifact. For example, if the object is deemed to be a user, biometric input may be acquired. The biometric input may then be processed to identify the user. The input device may be used at various locations, such as at an entry portal, point of sale, and so forth.
This disclosure is directed to, in part, mobile carts and/or associated computing devices that are configured to determine their respective location, determine an item in a facility that is associated with an item location that is nearest the location of the cart, determine one or more items that have been designated as similar to the nearest item, and output recommendation data corresponding to the similar items. In some instances, items may be ranked according to similarity based on a comparison of text data associated with the items, based on a comparison of location data associated with the items in other facilities, or the like.
B62B 3/14 - Hand carts having more than one axis carrying transport wheelsSteering devices thereforEquipment therefor characterised by provisions for nesting or stacking, e.g. shopping trolleys
In a data lake, a control data object is defined. The control object defines the processes and relationships of processes associated with a data set in the data lake. The control has states that are tied to and adapt in response to state changes of the associated data set. A control can have a control type. The system automatically carries forward enabled processes from one data set version to the next data set version. The system uses the control definition to execute processes, such as compaction or data quality scans, on data sets in the data lake.
Techniques for zero-shot and few-shot transfer of domain-adapted base networks are described. Multiple machine learning task layers are trained using a shared base feature extractor network. At least one task layer is trained with samples and corresponding labels from a first domain as well as a second domain. At least one other task layer is trained with samples and corresponding labels from only the first domain. Ultimately, the other task layer(s) are adapted to generate labels for the first domain via the base network being weighted based on all trainings.
This disclosure is directed to techniques in which a first user in an environment scans visual indicia associated with an item, such as a barcode, before handing the item to a second user. One or more computing devices may receive an indication of the scan, retrieve image data of the interaction from a camera within the environment, identify the user that received the item, and update a virtual cart associated with the second user to indicate addition of the item.
G06K 7/14 - Methods or arrangements for sensing record carriers by electromagnetic radiation, e.g. optical sensingMethods or arrangements for sensing record carriers by corpuscular radiation using light without selection of wavelength, e.g. sensing reflected white light
A vehicle data management system and data jurisdiction system manage vehicle data between multiple jurisdictions and enables a set of jurisdiction rules involving rules of various jurisdictions to be applied consistently. The vehicle data jurisdiction system can detect changes in jurisdiction of a vehicle based on various pieces of received vehicle information and applies appropriate jurisdiction rules from a set of jurisdiction rules. Various jurisdictions may have conflicting jurisdiction rules and, in such circumstances, the data jurisdiction system resolves potential conflicts between the rules using a jurisdiction rules resolution workflow. Based on the resolution of the conflict, the data jurisdiction system can migrate data of the vehicle to one or more other jurisdictions, or otherwise implement the correct rules determined by resolving the conflict.
Systems and methods are disclosed for containers having mechanical secured door indicators. In one embodiment, an example container may include a first container wall, a second container wall, a third container wall, and a fourth container wall configured to move from an open position to a closed positon. The fourth container wall may include a first portion coupled to the second container wall, and a second portion coupled to the third container wall, where the first portion and the second portion are coplanar when the fourth container wall is in the closed position. The foldable container may include a bottom container platform having a machine-readable code, and a pivotable arm configured to cover the machine-readable code when the fourth container wall is in the closed position, and to uncover the machine-readable code when the fourth container wall is not in the closed position.
B62B 5/00 - Accessories or details specially adapted for hand carts
B65D 6/18 - Containers having bodies formed by interconnecting or uniting two or more rigid, or substantially rigid, components made wholly or mainly of metal, plastics, wood or substitutes therefor collapsible with hinged components
24.
Techniques for identifying semantically relevant search results
Techniques are described herein for generating a semantically related search query that differs by at least one lexical token from document data provided as input. A machine-learning model may be trained using supervised and/or deep learning algorithms and a training data set including historical queries and the document data identified as being semantically related to those queries. The semantically-related search query may be associated with the document and utilized in subsequent searches to match the document to a subsequent query based on identifying a lexical match between the subsequent search query and the semantically-related search query associated with the document. These techniques enable semantically related search queries to be assigned offline but utilized to identify semantically related documents using lexical matching techniques. Identifying these types of matches create more diverse search results while maintaining a high degree of relevance between the query and the documents of the result set.
Maintaining the security of biometric data is an utmost priority. Biometric data is secured using one or more techniques. With one technique, biometric input such as images of a user's palm is used to generate first primary data (PD). The original biometric input is deleted from temporary secure storage while the first PD is securely stored. The first PD may then be processed later to determine a second PD. The first PD may then be deleted, and the second PD subsequently used. With another technique, biometric input or a PD may be processed by a first model to determine first secondary data (SD) that is representative of features of a particular user within a first embedding space. Later the PD may be processed by a second model to determine a second SD in a second embedding space. The first SD is deleted, and the second SD subsequently used.
Techniques for cooperative timing alignment using synchronization pulses are described. The techniques can include generating, at an integrated circuit device, a timing signal, controlling a local count value based on the timing signal, monitoring a synchronization signal of a system comprising the integrated circuit device, detecting a synchronization pulse in the synchronization signal, and aligning the local count value with an implied count value associated with the synchronization pulse in order to align the local count value with those of other integrated circuit devices of the system.
G06F 21/75 - Protecting specific internal or peripheral components, in which the protection of a component leads to protection of the entire computer to assure secure computing or processing of information by inhibiting the analysis of circuitry or operation, e.g. to counteract reverse engineering
27.
Reservation persistence in distributed block storage systems
A storage object and an associated permissions record is stored at a storage server. The permissions record indicates that some storage consumers are not permitted to perform a type of I/O operation on the storage object. In response to detecting that an event of a deletion triggering type with respect to the records, a modified version of the permissions record is stored at the server, indicating that the storage consumers remain prohibited from performing the I/O operations. In response to receiving a command to perform a particular I/O at the server after the modified version has been stored, the modified version is used to process the command.
A schema and cell value aware Named Entity Recognition (NER) model is used to perform natural language queries. Natural language queries may be received via an interface of a natural language query processing system. A fuzzy search may be performed that allows non-exact matches for column names or cell values of data sets potentially used to answer the natural language query. An NER model that adds a type embedding for an exact match of a column name or cell found in the fuzzy search that corresponds to a span of one or more words may be applied as part of generating the entity prediction for the natural language query. One or more queries to at least one of the data sets may be performed to return a result to the natural language query using the entity prediction generated by the NER machine learning model.
A technique to program a compute channel having multiple computational circuit blocks coupled in series in a pipeline can include receiving a machine instruction for the compute channel. The machine instruction is decoded to obtain an opcode, and the opcode can be used as an index to access an opcode entry in an opcode table. The opcode entry contains a pointer to a microoperation, and the pointer can be used to access a microoperation represented by a control entry in a control table and a datapath configuration entry in a datapath table. The microoperation can then be issued to the compute channel by configuring the compute channel with the control entry and the datapath configuration entry.
This disclosure describes a system for fulfilling items at a materials handling facility. In some instances, a predicted items list that identifies items that are likely to be picked by a user are determined and, when the user arrives at the materials handling facility, those predicted items are presented to the user for selection. For example, predicted items may be determined and an inventory holder that holds one or more of those predicted items may be retrieved by a mobile drive unit (such as a Kiva mobile drive unit) and presented to the user at a retrieval area. The user may pick the items they desire from the presented inventory holder.
Enhanced graphical user interfaces for transcription of audio and video messages is disclosed. Audio data may be transcribed, and the transcription may include emphasized words and/or punctuation corresponding to emphasis of user speech. Additionally, the transcription may be translated into a second language. A message spoken by a user depicted in one or more images of video data may also be transcribed and provided to one or more devices.
G10L 15/22 - Procedures used during a speech recognition process, e.g. man-machine dialog
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 3/0482 - Interaction with lists of selectable items, e.g. menus
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 40/109 - Font handlingTemporal or kinetic typography
G06F 40/40 - Processing or translation of natural language
G06F 40/47 - Machine-assisted translation, e.g. using translation memory
G06F 40/58 - Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation
G06V 40/20 - Movements or behaviour, e.g. gesture recognition
A system configured to improve audio processing by performing dereverberation and noise reduction during a communication session. In some examples, the system may include a deep neural network (DNN) configured to perform speech enhancement, which is located after an Acoustic Echo Cancellation (AEC) component. For example, the DNN may process isolated audio data output by the AEC component to jointly mitigate additive noise and reverberation. In other examples, the system may include a DNN configured to perform acoustic interference cancellation, which may jointly mitigate additive noise, reverberation, and residual echo, removing the need to perform residual echo suppression processing. The DNN is configured to process complex-valued spectrograms corresponding to the isolated audio data and/or estimated echo data generated by the AEC component.
G10L 21/10 - Transforming into visible information
G10L 25/30 - Speech or voice analysis techniques not restricted to a single one of groups characterised by the analysis technique using neural networks
33.
Performance characteristic transfer for localized content
Systems and techniques for validation and generation of localized content for audio and video are described herein. The systems and techniques provide for training of twin neural networks to evaluate performance characteristics, sometimes referred to as content-auxiliary characteristics, of a localized performance. The localized performance may be validated or improved by identifying misalignment in the performance characteristics to ensure that localized content preserves content as well as creative intent and performance ability in the final product. The machine learning models trained using the techniques described herein may be used in connection with auto-localization processes to automatically generate high quality localized audio and video content.
G11B 27/031 - Electronic editing of digitised analogue information signals, e.g. audio or video signals
G06F 40/58 - Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation
G06V 10/774 - Generating sets of training patternsBootstrap methods, e.g. bagging or boosting
G06V 10/82 - Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
G06V 20/40 - ScenesScene-specific elements in video content
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
G10L 17/04 - Training, enrolment or model building
G10L 17/06 - Decision making techniquesPattern matching strategies
G10L 25/57 - Speech or voice analysis techniques not restricted to a single one of groups specially adapted for particular use for comparison or discrimination for processing of video signals
34.
Adaptive testing service that generates test cases from observed behaviors
Techniques are disclosed to implement an adaptive testing service (ATS) capable of automatically generating test cases for a network service to adapt test coverage to observed behaviors of the network service. In embodiments, the ATS uses telemetry data from a production version of the network service to identify classes of testable behaviors. Test cases are generated for the behaviors and assigned weights based on frequency or recency metrics of the behaviors. The test cases are stored in a test case repository, and may be used to monitor the production version of the network service or verify code changes to a development version of the network service. The test case weights may be used to select which test cases to run or determine whether code changes should be accepted or rejected. The test cases are evolved over time to adapt to behavior changes in the network service.
Techniques for guiding camera focusing using barcode images are described herein. In an example, a computer system receives an image of a calibration target having a center that is aligned with an optical axis of a camera that generated the image. The calibration target includes steps that are located at different depths with respect to a lens of the camera. The calibration target includes barcode sets. Individual barcodes of a barcode set have a barcode attribute. The computer system decodes the barcode sets in the image and determines a set of decoded barcodes of the barcode sets. The computer system determines, for individual steps, a sharpness metric based on the barcode attribute of individual barcodes of the set of decoded barcodes and determines an adjustment of a focal plane of the lens based on the sharpness metric for the individual steps.
H04N 23/67 - Focus control based on electronic image sensor signals
G06K 7/14 - Methods or arrangements for sensing record carriers by electromagnetic radiation, e.g. optical sensingMethods or arrangements for sensing record carriers by corpuscular radiation using light without selection of wavelength, e.g. sensing reflected white light
A device includes a housing, a camera at least partially disposed within the housing, one or more microphones at least partially disposed within the housing, and a printed circuit board (PCB) including a switch. The switch has a lever that is transitionable between a displaced position and a non-displaced position. A privacy cover couple to the housing and includes a rib disposed on an interior surface of the privacy cover. The privacy cover is transitionable between a first position in which the camera is obstructed and the rib is engaged with the lever to transition the lever to the displaced position, and a second position in which the camera is unobstructed and the rib is disengaged with the lever such that the lever is permitted to transition to the non-displaced position.
The present disclosure generally relates to a first network device in a primary region that can failover network traffic into a second network device in a failover region. The first network device can receive routing criteria identifying how traffic originating in the primary region should be routed. The first network device can transmit this routing criteria to the second network device in the failover region. Based on determining the occurrence of a failover event, the first network device may transmit network traffic originating in the primary region to the second network device in the failover region. The second network device can determine how to route the network traffic based on the routing criteria of the primary region. In some embodiments, the second network device can determine how to route the network traffic based on the routing criteria of the failover region.
G06F 11/07 - Responding to the occurrence of a fault, e.g. fault tolerance
G06F 11/20 - Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements
Devices and techniques are generally described for key-value store having improved support for indexing and querying. The proposed system can expand query functionality by adding compound indexing. In some implementations, compound index may allow multiple simple key-value store queries (e.g., for multiple event types) to be replaced with a single query. In some implementations, compound index may allow for query filtering otherwise unavailable for the key-value store. The system may return results from the compound query in timestamp order and with appropriate pagination. Indexing may be updated dynamically; for example, manually via a self-service portal and/or automatically in response to frequent query combinations. The expanded query functionality can improve the efficiency and usability of, for example, an event timeline system.
An object data store may generate metadata responsive to a request that causes a scan of a data object for subsequent use in performing queries to the data object. A request may be received that causes a scan operation of the data object. As part of performing the scan one or multiple types of metadata describing the data object may be generated. The generated metadata may be applied to access the data object and perform a subsequently received query to the data object at the object data store.
Generated instruction sequences captured from software interactions may be executed as part of formal verification of a design under test. Software-instructed commands to be performed to configure a design under test formatted according to an interface implemented by the design under test can be obtained. A sequence to perform the software-instructed commands may be generated to configure the design under test in a hardware design and verification language. The sequence may then be executed to perform the software-instructed commands to configure the design under test and then perform formal verification on the configured design under test.
G06F 30/3323 - Design verification, e.g. functional simulation or model checking using formal methods, e.g. equivalence checking or property checking
G06F 30/367 - Design verification, e.g. using simulation, simulation program with integrated circuit emphasis [SPICE], direct methods or relaxation methods
43.
Systems for determining elevations of skin features
A wearable device may compress the skin of a user when worn, which may affect values determined using sensors of the device. To determine the effect of skin compression on the values, a time-of-flight signal, images, or frames of video data that depict a portion of the body having indentations from wearing the device may be acquired. Characteristics of the images, such as shadows associated with the indentations, may be processed using a machine learning algorithm or mathematical function to determine a depth of various portions of the indentations. Depth data from the time of flight signal may be used to refine or modify these determined depths. The amount of skin compression associated with the indentations may be used to modify signals acquired using sensors, or output a recommendation for a band or other method for securing the device.
Systems and techniques for optimizing deployment of a camera scanning system in an environment for item identification are described. An example technique involves obtaining a first set of parameters of the camera scanning system and obtaining a second set of parameters of the environment. A third set of parameters of a predicted scan volume of the camera scanning system are determined based on the first set of parameters and the second set of parameters. At least one of the first or second sets of parameters is modified upon determining that the predicted scan volume satisfies a first predetermined condition. An indication of at least one of the first, second, or third sets of parameters is transmitted upon determining that the predicted scan volume satisfies a second predetermined condition.
Systems, devices, and methods are described herein for improving object detection frameworks. Proposed regions can be used to identify similar images from a novel image set. Once identified, a weighted average of the feature representations of the similar images and/or a probability distribution of the classification labels for those images can be generated. The weighted average of the feature representations and/or the probability distribution can be used to steer the predicted classification confidence and/or predicted bounding box coordinates of the object detection framework. The disclosed techniques can be easily integrated with the object detect framework to improve the accuracy of its predictions without adding additional trainable parameters so as to refrain from adding complexity to the learning process.
G06V 10/774 - Generating sets of training patternsBootstrap methods, e.g. bagging or boosting
G06V 10/22 - Image preprocessing by selection of a specific region containing or referencing a patternLocating or processing of specific regions to guide the detection or recognition
G06V 10/74 - Image or video pattern matchingProximity measures in feature spaces
G06V 10/75 - Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video featuresCoarse-fine approaches, e.g. multi-scale approachesImage or video pattern matchingProximity measures in feature spaces using context analysisSelection of dictionaries
G06V 10/764 - Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
G06V 10/82 - Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
Real-time audio enhancement for a target speaker may be performed. An embedding of a sample of speaker audio is created using a trained neural network that performs voice identification. The embedding is then concatenated with the input features of a trained machine learning model for audio enhancement. The audio enhancement model can recognize and enhance a target speaker's speech in a real-time implementation, as the embedding is in the same feature space of the audio enhancement model.
Technologies directed to a hybrid-feed network of a parallel feed network and series-fed sub-arrays are described. The hybrid-feed network includes a parallel feed network and multiple groups of series-fed tiles, each tile including a beamforming integrated circuit (IC) and a set of antenna elements.
H01Q 3/40 - Arrangements for changing or varying the orientation or the shape of the directional pattern of the waves radiated from an antenna or antenna system varying the relative phase or relative amplitude of energisation between two or more active radiating elementsArrangements for changing or varying the orientation or the shape of the directional pattern of the waves radiated from an antenna or antenna system varying the distribution of energy across a radiating aperture varying the phase by electrical means with phasing matrix
H01Q 3/26 - Arrangements for changing or varying the orientation or the shape of the directional pattern of the waves radiated from an antenna or antenna system varying the relative phase or relative amplitude of energisation between two or more active radiating elementsArrangements for changing or varying the orientation or the shape of the directional pattern of the waves radiated from an antenna or antenna system varying the distribution of energy across a radiating aperture
H01Q 3/36 - Arrangements for changing or varying the orientation or the shape of the directional pattern of the waves radiated from an antenna or antenna system varying the relative phase or relative amplitude of energisation between two or more active radiating elementsArrangements for changing or varying the orientation or the shape of the directional pattern of the waves radiated from an antenna or antenna system varying the distribution of energy across a radiating aperture varying the phase by electrical means with variable phase-shifters
H01Q 5/35 - Individual or coupled radiating elements, each element being fed in an unspecified way for different propagation modes using two or more simultaneously fed points
Techniques are disclosed for migrating a computer application from an entity's premises to a web services platform. Data from multiple sources on the entity's premises is gathered and normalized into a common format. The normalized data is used to create a topology of the network on the entity's premises. This topology is analyzed to determine whether a computer application executing on the entity's premises may be migrated to the web service platform.
Media content may be mastered at a higher quality than is supported on various remote workstations to perform tasks with respect to that content, using transmission channels that do not support data transfer rates for large, high quality media content. A compressed version of this content may be transmitted over a first channel for use with various tasks on a remote workstation. For tasks such as color grading that benefit from this higher quality content, a separate but parallel communication channel is used to transmit a higher-quality version of this content. An uncompressed video stream can be encoded using a lossless codec to retain higher quality data. A high quality video stream can be transmitted over a separate transmission channel, and received to a decoder that can decode this stream to provide a high quality video signal for presentation via a grading monitor or other such high quality presentation device.
H04N 19/186 - 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 a colour or a chrominance component
H04N 19/184 - 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 bits, e.g. of the compressed video stream
50.
Establishing communication links to assist headless devices
Techniques for establishing connections between user devices and headless devices attempting to connect to networks. A headless device may attempt to connect to an access point that requires interaction with a captive portal webpage for access to a network. However, the headless device my lack a display to present the captive portal webpage. The headless device may establish a connection with a user device using a PAN protocol. The headless device may then receive the captive portal webpage received from the access point, and relay the webpage to the user device using the PAN protocol. A user may use the user device to interact with the captive portal webpage, and the user device may then relay interaction data back to the headless device using the PAN protocol. The headless device may then provide that interaction data to the access point to be provided access to the network.
H04W 4/80 - Services using short range communication, e.g. near-field communication [NFC], radio-frequency identification [RFID] or low energy communication
Generative pre-trained large language models (LLMs) can create domain-specific text answers in various formats like JSON, XML, HTML, SQL, or programming languages. However, LLMs may “hallucinate,” generating incorrect or nonsensical answers that diverge from reality, thus eroding trust in their outputs or worse. Disclosed techniques use a sampling-based approach and an equivalence checker. Multiple answers (samples) to a prompt are generated by the LLM; if they are equivalent, the LLM is likely answering correctly. If the samples disagree or contradict, it's more likely that the LLM is hallucinating, or the prompt is ambiguous. An automated reasoning equivalence checker is utilized to verify the samples' functional equivalency, providing a method to detect and possibly rectify hallucination issues in LLM-generated answers.
Systems and methods are provided for an on-demand code execution service comprising a set of computing devices for on-demand execution of function code while continuing to facilitate executing long-running background processes. A subset of resources may be initialized based, at least in part, on the application configuration data including at least a request-response process, a background process, and a lesser set of computing resources for the background process. After the execution of the background process has begun, a first request may be received. The on-demand code execution service may increase computing resources to a larger set of computing resources to generate a first response to the first request. The first response may then be provided to an external set of computing resources. After determining that the queue contains no additional requests, the on-demand code execution service may decrease the level of computing resources to the lesser set of computing resources.
Distributed orchestration of data retrieval for generative machine learning model may be performed. When a natural language request to perform a natural language task is received that is associated with a generative application, one or more data retrievers may be selected to access associated data repositories according to a previously specified retrieval configuration for the generative natural language application. The data may then be obtained by the selected data retrievers and used to generate a prompt to a generative machine learning model. A result of the generative machine learning model may then be used to provide a response to the natural language request to perform the natural language task.
Intent classification is performed for executing a retrieval augmented generation pipeline for natural language tasks using a generative machine learning model. A natural language generative application with associated data repositories may submit a natural language task. A classification machine learning model is used to determine an intent for the natural language request. A number of iterations of a retrieval pipeline may be determined to perform the natural language task based on the intent. The natural language request may be processed through a retrieval pipeline according to the determined number of iterations before returning a result to the request.
Template-based tuning is performed on a generative machine learning model where a shared template is used to tune the generative machine learning model across multiple natural language tasks. When a natural language request to perform a natural language task is received, portions of a shared template to complete are identified as part of generating a prompt. The generative machine learning model is instructed according to the generated prompt and a response to the request is returned based on a result of the generative machine learning model.
Techniques for a knowledge-graph system to use large language models (LLMs) to build knowledge graphs to answer queries submitted to a chatbot by users. The knowledge-graph system builds the knowledge graph using answers produced by an ELM for novel queries. The chatbot will continue to use the ELM to answer novel queries, but the chatbot may harness the knowledge graph to answer repeat questions to gain various efficiencies over LLM- backed chatbots. For example, the knowledge-graph system may easily debug or otherwise improve the answers in knowledge graphs, store provenance information in knowledge graphs, and augment the knowledge graphs using other data sources. Thus, the reliability and correctness of chatbots will be improved as the bugs and inaccuracies in answers provided by the ELM will be corrected in the knowledge graphs, but the chatbots can still harness the abilities of LLMs to provide answers across various subject-matter domains.
A system (100; 200; 500-700) and supporting method (300, 400) enable receipt (302) of a computer-coded policy for execution in a control plane associated with a cloud environment to provide data governance in a data plane using one or more data assets of the cloud environment, where the one or more data assets are automatically associated (310) to the computer-coded policy using a set of pre-determined rules associated with the computer-coded policy and using annotations associated with the one or more data assets, and where dynamic changes (404) are to be performed with respect to the annotations based in part on real-time changes to the computer-coded policy to allow monitoring (406) contents of the one or more data assets in accordance with the computer-coded policy and to perform (412) a remediation action that is associated with the one or more data assets in response to a violation associated with the computer-coded policy.
Systems and methods are provided for managing computing services for an application comprising a plurality of virtual computing components executing on one or more host computing devices, wherein a service virtual computing component is to perform application functionality, and wherein a system computing component is to perform system functionality including management of the application virtual computing component; determining the service virtual computing component is to execute using a first access credential to provide a first computing service to the application virtual computing component, and the service virtual computing component is to execute using a second access credential to provide a second computing service to the system computing component, wherein the first access credential is assigned a different set of computing resource access permissions than the second access credential.
Systems and methods are provided for an on-demand code execution service comprising a set of computing devices for on-demand execution of function code while continuing to facilitate executing long-running background processes. A subset of resources may be initialized based, at least in part, on the application configuration data including at least a request-response process, a background process, and a lesser set of computing resources for the background process. After the execution of the background process has begun, a first request may be received. The on-demand code execution service may increase computing resources to a larger set of computing resources to generate a first response to the first request. The first response may then be provided to an external set of computing resources. After determining that the queue contains no additional requests, the on-demand code execution service may decrease the level of computing resources to the lesser set of computing resources.
Generative pre-trained large language models (LLMs) can create domain-specific text answers in various formats like JSON, XML, HTML, SQL, or programming languages. However, LLMs may "hallucinate," generating incorrect or nonsensical answers that diverge from reality, thus eroding trust in their outputs or worse. Disclosed techniques use a sampling-based approach and an equivalence checker. Multiple answers (samples) to a prompt are generated by the LLM; if they are equivalent, the LLM is likely answering correctly. If the samples disagree or contradict, it's more likely that the LLM is hallucinating, or the prompt is ambiguous. An automated reasoning equivalence checker is utilized to verify the samples' functional equivalency, providing a method to detect and possibly rectify hallucination issues in LLM-generated answers.
Examples herein provide an approach to enhance an audio mixture of a teleconference application by switching between noise suppression modes using a single model. Specifically, a machine learning (ML) model may be configured to, in response to receiving an audio mixture representation as input, suppress either a background noise of the audio mixture or suppress all noise of the audio mixture except a user's voice. In some examples, the ML model may be trained on speech and background noise training data during a training phase. In addition, the ML model may be trained on a user's voice during an enrollment phase. In addition, during an inference phase, the ML model may enhance the audio mixture by suppressing a portion of the audio mixture.
G10L 25/30 - Speech or voice analysis techniques not restricted to a single one of groups characterised by the analysis technique using neural networks
H04M 3/56 - Arrangements for connecting several subscribers to a common circuit, i.e. affording conference facilities
A plurality of location indications may be received that indicate locations of an object at a plurality of times. A plurality of geofence indications that indicate that the object is within a geofence may be generated based on the plurality of location indications. A plurality of notifications of a plurality of geofence events corresponding to the object may be provided, to an account, based on the plurality of geofence indications. The plurality of geofence events may include a geofence entering event and a geofence exiting event. An out-of-order location indication may be detected within the plurality of location indications. A retroactive geofence event regarding which the account has not yet been notified may be determined based on the out-of-order location indication. An additional notification of the retroactive geofence event may be provided to the account.
Techniques for caching in a machine learning model (ML) hosting service are described. ML model usage data is aggregated from host usage data provided from each host of a first set of hosts, the ML model usage data including, for a particular ML model, a number of inference requests to the particular ML model. A priority order of hosts in a second set of hosts to service an inference request for the particular ML model is calculated. Based on the ML model usage data and the priority order, a set of ML models to load to a particular host in the second set of hosts is determined. The particular host is caused to load the set of ML models. A router is updated to direct ML model inference requests amongst the second set of hosts.
An index is created with split documents to retrieve and augment generation of a response to a natural language request using a generative machine learning model. When a natural language request is received, a search representation is generated and used to retrieve candidate portions of documents from the index. A relevancy ranking is performed to identify relevant portions of documents from the candidates and provide the relevant portions to prompt a generative machine learning model to provide a result for the natural language request.
Distributed orchestration of data retrieval for generative machine learning model may be performed. When a natural language request to perform a natural language task is received that is associated with a generative application, one or more data retrievers may be selected to access associated data repositories according to a previously specified retrieval configuration for the generative natural language application. The data may then be obtained by the selected data retrievers and used to generate a prompt to a generative machine learning model. A result of the generative machine learning model may then be used to provide a response to the natural language request to perform the natural language task.
A set of alternative vocal input styles for specifying a parameter of a dialog-driven application is determined. During execution of the application, an audio prompt requesting input in one of the styles is presented. A value of the parameter is determined by applying a collection of analysis tools to vocal input obtained after the prompt is presented. A task of the application is initiated using the value.
An index is created with split documents to retrieve and augment generation of a response to a natural language request using a generative machine learning model. When a natural language request is received, a search representation is generated and used to retrieve candidate portions of documents from the index. A relevancy ranking is performed to identify relevant portions of documents from the candidates and provide the relevant portions to prompt a generative machine learning model to provide a result for the natural language request.
Techniques for caching in a machine learning model (ML) hosting service are described. ML model usage data is aggregated from host usage data provided from each host of a first set of hosts, the ML model usage data including, for a particular ML model, a number of inference requests to the particular ML model. A priority order of hosts in a second set of hosts to service an inference request for the particular ML model is calculated. Based on the ML model usage data and the priority order, a set of ML models to load to a particular host in the second set of hosts is determined. The particular host is caused to load the set of ML models. A router is updated to direct ML model inference requests amongst the second set of hosts.
Systems and methods are provided for managing computing services for an application comprising a plurality of virtual computing components executing on one or more host computing devices, wherein a service virtual computing component is to perform application functionality, and wherein a system computing component is to perform system functionality including management of the application virtual computing component; determining the service virtual computing component is to execute using a first access credential to provide a first computing service to the application virtual computing component, and the service virtual computing component is to execute using a second access credential to provide a second computing service to the system computing component, wherein the first access credential is assigned a different set of computing resource access permissions than the second access credential.
Techniques for a knowledge-graph system to use large language models (LLMs) to build knowledge graphs to answer queries submitted to a chatbot by users. The knowledge-graph system builds the knowledge graph using answers produced by an LLM for novel queries. The chatbot will continue to use the LLM to answer novel queries, but the chatbot may harness the knowledge graph to answer repeat questions to gain various efficiencies over LLM-backed chatbots. For example, the knowledge-graph system may easily debug or otherwise improve the answers in knowledge graphs, store provenance information in knowledge graphs, and augment the knowledge graphs using other data sources. Thus, the reliability and correctness of chatbots will be improved as the bugs and inaccuracies in answers provided by the LLM will be corrected in the knowledge graphs, but the chatbots can still harness the abilities of LLMs to provide answers across various subject-matter domains.
H04L 51/02 - User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages
Examples herein provide an approach to enhance an audio mixture of a teleconference application by switching between noise suppression modes using a single model. Specifically, a machine learning (ML) model may be configured to, in response to receiving an audio mixture representation as input, suppress either a background noise of the audio mixture or suppress all noise of the audio mixture except a user's voice. In some examples, the ML model may be trained on speech and background noise training data during a training phase. In addition, the ML model may be trained on a user's voice during an enrollment phase. In addition, during an inference phase, the ML model may enhance the audio mixture by suppressing a portion of the audio mixture.
G10L 25/30 - Speech or voice analysis techniques not restricted to a single one of groups characterised by the analysis technique using neural networks
H04M 3/56 - Arrangements for connecting several subscribers to a common circuit, i.e. affording conference facilities
88.
RELIABLE PACKET DELIVERY THROUGH PATH DIVERSIFICATION
Approaches are disclosed for providing path diversity in a data transmission network. A primary transmission path can be selected through a network, such as a backbone network, based on factors such as cost of transmission. At least one waypoint can be selected that is to be included in a secondary transmission path. The waypoint(s) can be selected such that the secondary transmission path will have few, if any, network components in common with the primary transmission path, providing significant path diversity. The waypoint(s) can be selected based on a cost ratio or other such factor. In the event of a failure of transmission of a data packet over one of the transmission paths, a second transmission attempt can be performed using the same path or the other transmission path, or both.
A plurality of location indications may be received that indicate locations of an object at a plurality of times. A plurality of geofence indications that indicate that the object is within a geofence may be generated based on the plurality of location indications. A plurality of notifications of a plurality of geofence events corresponding to the object may be provided, to an account, based on the plurality of geofence indications. The plurality of geofence events may include a geofence entering event and a geofence exiting event. An out-of-order location indication may be detected within the plurality of location indications. A retroactive geofence event regarding which the account has not yet been notified may be determined based on the out-of-order location indication. An additional notification of the retroactive geofence event may be provided to the account.
Techniques for a knowledge-graph system to use large language models (LLMs) to build knowledge graphs to answer queries submitted to a chatbot by users. The knowledge-graph system builds the knowledge graph using answers produced by an LLM for novel queries. The chatbot will continue to use the LLM to answer novel queries, but the chatbot may harness the knowledge graph to answer repeat questions to gain various efficiencies over LLM-backed chatbots. For example, the knowledge-graph system may easily debug or otherwise improve the answers in knowledge graphs, store provenance information in knowledge graphs, and augment the knowledge graphs using other data sources. Thus, the reliability and correctness of chatbots will be improved as the bugs and inaccuracies in answers provided by the LLM will be corrected in the knowledge graphs, but the chatbots can still harness the abilities of LLMs to provide answers across various subject-matter domains.
G06N 3/042 - Knowledge-based neural networksLogical representations of neural networks
G06N 3/006 - Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
Techniques for using a language model (e.g., a large language model (LLM)) to generate a natural language response to a user input and prosody information (e.g., voice characteristics associated with a synthetic voice to output the natural language response to the user) are described. The prosody information may correspond to a natural language (e.g., text or tokenized) description, a spectrogram, and/or a latent representation of the voice characteristic(s) associated with the natural language response. In some embodiments, the natural language response and the prosody information may be generated by different portions of layers of the language model. In such embodiments, the output of the layer(s) of the language model configured to generate the natural language response may be provided to the layer(s) of the language model configured to generate the prosody information and the output may be used to generate the prosody information, and vice versa.
Techniques for unlearning concepts in the use of a pre-trained generative machine learning model are described. A description of a concept to be unlearned in use of a pre-trained generative machine learning model is received. Negative prompts and positive prompts are processed with the pre-trained generative machine learning model to generate associated activation volume maps. A set of conditions to differentiate activation volume maps associated with negative prompts from activation volume maps associated with positive prompts is identified. A model adapter is generated, the model adapter to use a set of different model parameters when processing of a prompt by the pre-trained generative machine learning model satisfies the set of conditions.
Systems and methods are disclosed for foldable containers with improved mobility. In one embodiment, an example foldable container may include a first container wall having a frame and a panel coupled to the frame, a second container wall coupled to the first container wall, the second container wall configured to rotate outwards with respect to the first container wall, a third container wall coupled to the first container wall, and a fourth container wall. The fourth container wall may include a first portion coupled to the second container wall, where the first portion rotates with respect to the second container wall, and a second portion coupled to the third container wall. The foldable container may include a latch configured to secure the first portion to the second portion, where the latch has a handle and a spring that biases that handle towards a front of the foldable container.
B65D 6/18 - Containers having bodies formed by interconnecting or uniting two or more rigid, or substantially rigid, components made wholly or mainly of metal, plastics, wood or substitutes therefor collapsible with hinged components
B62B 3/02 - Hand carts having more than one axis carrying transport wheelsSteering devices thereforEquipment therefor involving parts being adjustable, collapsible, attachable, detachable, or convertible
B62B 5/06 - Hand moving equipment, e.g. handle bars
Systems and methods are disclosed for structural rib assemblies for container shuttle rails. In one embodiment, an example system for a shuttle may include a rib assembly. The rib assembly may include a first plate with a first tab on a first lateral side, a second tab on a second lateral side, a first cutout disposed adjacent to the first tab, a second cutout disposed adjacent to the second tab, and a third cutout disposed along an upper portion of the first plate. The rib assembly may include a second plate coupled to the first plate, the second plate having a third tab on a first lateral side, a fourth tab on a second lateral side, a fourth cutout disposed adjacent to the third tab, a fifth cutout disposed adjacent to the fourth tab, and a sixth cutout disposed along an upper portion of the second plate.
B65G 21/02 - Supporting or protective framework or housings for endless load-carriers or traction elements of belt or chain conveyors consisting essentially of struts, ties, or like structural elements
B65G 54/02 - Non-mechanical conveyors not otherwise provided for electrostatic, electric, or magnetic
Techniques for handling natural language query processing are described. In some examples, a sequence-to-sequence model is used to handle a natural language query. Post-processing of a result of the sequence-to-sequence model utilizes fine-grained information from an entity linker. In some examples, the sequence-to-sequence model and aspects of a natural language query pipeline are used to handle a natural language query.
Captures or recordings of sensitive information or data displayed on screens or displays are detected by generating unique identifiers of users and embedding linked codes including such identifiers into the information or data. When the information or data is accessed by a user and displayed on a screen, and an image of the information or data is captured by a camera of a mobile device or other system, the camera detects a code within the images and requests to access a page or other networked resource associated with a link embedded in the code. Upon detecting a request to access such a page, the request may be attributed to the user. Upon detecting a unique identifier within an image depicting sensitive information, the image may be attributed to the user.
