Abstract: A method for classifying a human-object interaction includes identifying a human-object interaction in the input. Context features of the input are identified. Each identified context feature is compared with the identified human-object interaction. An importance of the identified context feature is determined for the identified human-object interaction. The context feature is fused with the identified human-object interaction when the importance is greater than a threshold.

Abstract: A method for managing model updates by a first zone server, associated with a first zone model of a plurality of zone models, includes receiving a global model from a global server associated with the global model. The method also includes transmitting the global model to user equipment (UEs) in a first group of UEs associated with the first zone model. The method further includes receiving, from one or more UEs in the first group, model updates associated with the global model based on transmitting the global model. The method further includes transmitting, to the global server, an average of the model updates received from the one or more UEs. The method also includes updating the global model to generate the first zone model based on the model updates. The method further includes transmitting the first zone model to one or more UEs in the first group.

Abstract: Techniques for biometric authentication using a voice accelerometer and an acoustic transducer includes receiving a first signal representing an output by a voice accelerometer of a device and a second signal representing an output by an acoustic transducer of the device. Characteristics of each of the first signal and the second signal are compared with stored biometric characteristics for a user of the device. The user of the device is then authenticated based on the comparison.

Abstract: A computer-implemented method for tracking with visual object constraints includes receiving a lingual constraint and a video. A word embedding is generated based on the lingual constraint. A set of features is extracted for one or more frames of the video. The word embedding is cross-correlated to the set of features for the one or more frames of the video. A prediction indicating whether the lingual constraint is in the one or more frames of the video is generated based on the cross-correlation.

Abstract: An electrochemical cell for sensing gas has added mechanical support for the working electrode to prevent flexure of the working electrode due to pressure differentials. The added mechanical support includes: 1) affixing a larger area of the working electrode to the body of the cell; 2) a gas vent to a cavity of the body to equalize pressures; 3) a rigid electrolyte layer abutting a back surface of the working electrode; 4) infusing an adhesive deep into sides of the porous working electrode to enhance rigidity; 5) supporting opposing surfaces of the working electrode with the rigid package body; and 6) other techniques to make the working electrode more rigid. A bias circuit is also described that uses a controllable current source, an integrator of the varying current, and a feedback circuit for supplying a voltage to the counter electrode and a bias voltage to the reference electrode.

Abstract: Certain aspects of the present disclosure provide a method for performing machine learning, comprising: determining a plurality of vertices in a neighborhood associated with a mesh including a target vertex; determining a linear transformation configured to parallel transport signals along all edges in the mesh to the target vertex; applying the linear transformation to the plurality of vertices in the neighborhood to form a combined signal at the target vertex; determining a set of basis filters; linearly combining the basis filters using a set of learned parameters to form a gauge equivariant convolution filter, wherein the gauge equivariant convolution filter is constrained to maintain gauge equivariance; applying the gauge equivariant convolution filter to the combined signal to form an intermediate output; and applying a nonlinearity to the intermediate output to form a convolution output.

Abstract: A computer-implemented method for contrastive object representation from temporal data using an artificial neural network (ANN) includes receiving, by the ANN, a video. The video comprises a temporal sequence of frames including images of one or more objects. The ANN generates object representations corresponding to the one or more objects based on temporal data of multiple frames of the temporal sequence of frames. The object representations are communicated to a receiver.

Abstract: A package that includes a substrate, an integrated device coupled to the substrate, an encapsulation layer located over the substrate, at least one encapsulation layer interconnect located in the encapsulation layer, and a metal layer located over the encapsulation layer. The substrate includes at least one dielectric layer and a plurality of interconnects. The encapsulation layer interconnect is coupled to the substrate. The metal layer is configured as an electromagnetic interference (EMI) shield for the package. The metal layer is located over a backside of the integrated device.

Abstract: A microphone including a casing having a front wall, a back wall, and a side wall joining the front wall to the back wall, a transducer mounted to the front wall, the transducer including a substrate and a transducing element, the transducing element having a transducer acoustic compliance dependent on the transducing element dimensions, a back cavity cooperatively defined between the back wall, the side wall, and the transducer, the back cavity having a back cavity acoustic compliance. The transducing element is dimensioned such that the transducing element length matches a predetermined resonant frequency and the transducing element width, thickness, and elasticity produces a transducer acoustic compliance within a given range of the back cavity acoustic compliance.

Abstract: An acoustic device is described and includes an acoustic sensor element configured to sense acoustic energy and produce an output signal and a threshold detector circuit including a switch having an input coupled to the output of the acoustic sensor element to receive the output signal, a control port that receives a control signal, and first and second output ports, a first channel including an analog-to-digital converter that operates at a first power level a second analog-to-digital converter that operates at a second higher power level, relative to the first power level and a threshold level detector that receives an output from the first analog-to-digital converter to produce the control signal having a first state that causes the switch feed the output signal from the acoustic sensor element to the second analog-to-digital converter when the first digitized output signal meets a threshold criteria.

Abstract: A transducer of the preferred embodiment including a transducer and a plurality of adjacent, tapered cantilevered beams. Each of the beams define a beam base, a beam tip, and a beam body disposed between the beam base and the beam tip. The beams are arranged such that each of the beam tips extends toward a common area. Each beam is joined to the substrate along the beam base and is free from the substrate along the beam body. A preferred method of manufacturing a transducer can include: depositing alternating layers of piezoelectric and electrode onto the substrate in block, processing the deposited layers to define cantilever geometry in block, depositing metal traces in block, and releasing the cantilevered beams from the substrate in block.

Abstract: A device that includes an adaptive acoustic detection circuit and an acoustic sensor device such as a microphone is described. The device includes in addition to the sensor a circuit configured to detect when an input stimulus to the sensor satisfies an adaptive threshold, and further configured to produce a signal upon detection that causes adjustment of performance of the device, wherein the adaptive threshold is a threshold value that varies over time in accordance with detected changes to sound of an environment in which the device is located.

Abstract: UE location determined by collecting and preprocessing signal data at a detector and sending extracted data to a remote locate server. The detector buffers samples from signals provided by receive channels, detects known reference signals from receive channels based on reference signal parameters, isolates symbols carrying the reference signal from frames, extracts data from symbols, and sends extracted data to locate server. The locate server receives the extracted data, estimates locate observables based on the extracted data and calculates the UE location based on the estimated locate observables, the reference signal parameters and the extracted data. The detector and/or the server may also generate correlation coefficients between reference signals carrying spectrum received from a serving cell and utilize the correlation coefficients to cancel a serving cell signal in symbols that include known in advance reference signals from the serving cell and one or more neighboring cells of the wireless system.

Abstract: An apparatus for testing a transducer module includes a test signal generator coupled to a common-mode terminal common to a plurality of transducers, and a signal processing circuit configured to receive output signal from each of said transducers and to produce an output signal. If the transducers are well matched to one another, the output signal will have little or no output amplitude. If there is a mismatch between the transducers, however, the output signal will have an amplitude proportional to the mismatch. The amplitude of the output signal may be compared to a predetermined threshold in order to produce a mismatch output signal indicating the existence of, and/or the degree of, mismatch between the transducers.

Abstract: A sensor device that senses proper acceleration. The sensor device includes a substrate, a spacer layer supported over a first surface of the substrate, at least a first tapered cantilever beam element having a base and a tip, the base attached to the spacer layer, and which is supported over and spaced from the substrate by the spacer layer. The at least first tapered cantilever beam element tapers in width from the base portion to the tip portion. The at least first cantilever beam element further including at least a first layer comprised of a piezoelectric material, a pair of electrically conductive layers disposed on opposing surfaces of the first layer, and a mass supported at the tip portion of the at least first tapered cantilever beam element.

Abstract: A sensor device that senses proper acceleration. The sensor device includes a substrate, a spacer layer supported over a first surface of the substrate, at least a first tapered cantilever beam element having a base and a tip, the base attached to the spacer layer, and which is supported over and spaced from the substrate by the spacer layer. The at least first tapered cantilever beam element tapers in width from the base portion to the tip portion. The at least first cantilever beam element further including at least a first layer comprised of a piezoelectric material, a pair of electrically conductive layers disposed on opposing surfaces of the first layer, and a mass supported at the tip portion of the at least first tapered cantilever beam element.

Abstract: Techniques for automatically muting or unmuting an acoustic transducer include receiving a signal representing an output by a voice accelerometer of a device, determining whether the signal is indicative of a presence or absence of voice activity by a user of the device, and generating a control signal that causes an acoustic transducer to be muted responsive to determining that the signal is indicative of an absence of voice activity by the user, or that causes the acoustic transducer to be unmuted response to determining that the signal is indicative of a presence of voice activity by the user.

Abstract: A transducer comprising: at least one piezoelectric layer; a first patterned conductive layer that is patterned with a first opening; a second patterned conductive layer that is patterned with a second opening; wherein at least one piezoelectric layer is between the first and the second patterned conductive layers in a stack; and wherein a position of the first opening is staggered relative to a position of the second opening in the stack to mitigate an occurrence of crack propagation through the layers.

Abstract: Certain aspects provide a method for determining a solution to a combinatorial optimization problem, including: determining a plurality of subgraphs, wherein each subgraph of the plurality of subgraphs corresponds to a combinatorial variable of the plurality of combinatorial variables; determining a combinatorial graph based on the plurality of subgraphs; determining evaluation data comprising a set of vertices in the combinatorial graph and evaluations on the set of vertices; fitting a Gaussian process to the evaluation data; determining an acquisition function for vertices in the combinatorial graph using a predictive mean and a predictive variance from the fitted Gaussian process; optimizing the acquisition function on the combinatorial graph to determine a next vertex to evaluate; evaluating the next vertex; updating the evaluation data with a tuple of the next vertex and its evaluation; and determining a solution to the problem, wherein the solution comprises a vertex of the combinatorial graph.

Abstract: Systems and methods for determining a location of one or more user equipment (UE) in a wireless system can comprise receiving reference signals via a location management unit having two or more co-located channels, wherein the two or more co-located channels are tightly synchronized with each other and utilizing the received reference signals to calculate a location of at least one UE among the one or more UE. Embodiments include multichannel synchronization with a standard deviation of less than or equal 10 ns. Embodiments can include two LMUs, with each LMU having internal synchronization, or one LMU with tightly synchronized signals.