Abstract: This disclosure provides implementations for updating a neural network to improve execution of the neural network, including reducing the number of operations to be executed and processing resources and time required to execute the neural network. The implementations also relate to updating the neural network without requiring changes to a neural network framework or firmware. In some aspects, updating the neural network includes replacing a defined object in the neural network (such as an operator or group of operators in a layer all the way up to the layer itself of the neural network) with a specialized operator that is to call a specialized kernel that optimizes the replaced object. The specialized kernel requires fewer computing resources (such as fewer operations) to be executed than the replaced object. The specialized kernel may replace a layer of the neural network or be inserted between layers of the neural network.

Abstract: Systems and methods for touch sensing are provided. An input device includes a display and a touch sensor. The touch sensor has a sensing area with a plurality of sensor pixels forming rows and columns. Each sensor pixel includes a first electrode having first capacitive coupling areas and a second electrode having second capacitive coupling areas. The first capacitive coupling areas are configured to capacitively couple with one or more adjacent sensor pixels disposed in a same row and the second capacitive coupling areas are configured to capacitively couple with one or more adjacent sensor pixels in a same column. The first electrode is configured to electrically connect to one or more sensor pixels in the same column and the second electrode is configured to electrically connect to one or more sensor pixels in the same row. The sensor pixels are configured to facilitate parallel touch sensing in multiple orientations.

Abstract: A system includes sensor electrodes, a touch controller, and a transmitter that provides a transmitter signal of a first frequency to an electrode capacitively coupled to a given object. The touch controller is configured to: generate a sensing signal that has a second frequency different from the first frequency during a touch sensing phase and has a DC waveform during a touch identification phase; generate touch sensing signal data based on the sensing signal with the second frequency and resulting signals received from the sensor electrodes during the touch sensing phase; sense a touch object based on the touch sensing signal data; generate touch identification signal data based on the sensing signal with the DC waveform and the resulting signals received from the sensor electrodes during the touch identification phase; and identify whether the sensed touch object is associated with the given object based on the touch identification signal data.

Abstract: This disclosure provides methods, devices, and systems for audio signal processing. The present implementations more specifically relate to speech enhancement techniques that utilize multi-channel audio signals for audio source classification. In some aspects, a speech enhancement system may include an adaptive filter, a feature extractor, and a feature classifier. The adaptive filter is configured to receive a multi-channel audio signal, via at least a first microphone and a second microphone, and determine a relative impulse response (ReIR) between the microphones based on the multi-channel audio signal. The feature extractor is configured to extract a set of features from the ReIR based at least in part on a peak of the ReIR. The feature classifier is configured to classify the set of features as being associated with a target source or a distractor source based on a Gaussian mixture model (GMM).

Abstract: This disclosure provides methods, devices, and systems for wireless communications. The present implementations more specifically relate to reducing the power consumption of radio frequency (RF) power amplifiers without sacrificing power efficiency. In some aspects, an RF transmitter may include a signal generator, a power amplifier, and a duty cycle controller. The signal generator is configured to produce one or more RF signals based on a modulation scheme, and the power amplifier is configured to amplify the one or more RF signals for transmission over a wireless communication channel. In some implementations, the duty cycle controller may adjust a duty cycle of each RF signal based on data to be transmitted according to the modulation scheme. By changing the duty cycle of the RF signal, the duty cycle controller may toggle the output power of the power amplifier and thus modulate the amplitude of the output waveform to carry the data.

Abstract: A docking station is operable in a plurality of modes. The docking station obtains first data via a first interface configured to communicate with a computing device when operating in a first mode. The docking station obtains second data via a second interface when operating in a second mode. The docking station selectively outputs to a display via a third interface the first data when operating in the first mode, the second data when operating in the second mode, or the second data overlaid on at least a portion of the first data in response to the docking station operating in a third mode. The docking station determines when the computing device is not coupled to the docking station and, in response, operates in the second mode. The second data may be on-screen display data for an on-screen menu obtained from a processor associated with the docking station.

Abstract: The present disclosure relates to a method of managing display data in a system comprising a host device and a plurality of client devices connected to the host device. The method comprises receiving display data from at least one client device of the plurality of client devices at the host device via a shared connection between the plurality of client devices and the host device; generating, at the host device, control information for controlling generation, encoding and/or transmission of the display data by the at least one client device of the plurality of client devices; and transmitting the control information from the host device to the at least one client device of the plurality of client devices.

Abstract: A display device includes a plurality of light sources, a plurality of display driver integrated circuits (DDICs), and a light source driver. The light sources are configured to illuminate a display panel comprising a plurality of regions. Each DDIC is configured to generate backlighting data for a respective region of the plurality of regions. The backlighting data is indicative of luminance levels of respective light sources, of the plurality of light sources, which correspond to the respective region of the plurality of regions. Each DDIC is further configured to store ordering information indicative of an order for outputting the backlighting data, and output the backlighting data for the respective region based on the ordering information and a backlighting data synchronization signal. The light source driver is configured to drive the plurality of light sources based on the backlighting data for the plurality of regions output from the plurality of DDICs.

Abstract: A method performed by a docking station operable in a plurality of modes is disclosed. The method may include obtaining first data via a first interface of the docking station and second data via a second interface of the docking station, responsive to operating in a first mode of the plurality of modes. The first interface may be configured to couple the docking station to a computing device, and the second interface may be configured to communicate with a network. The method may also include obtaining third data via the second interface of the docking station, in lieu of the first interface, responsive to operating in a second mode of the plurality of modes. The method may further include selectively outputting the first data and the second data, or the third data, to a display based on whether the docking station operates in the first mode or the second mode.

Abstract: This disclosure provides methods, devices, and systems for exposure control for digital images. The present implementations more specifically relate to dynamic region-of-interest (ROI) based auto-exposure control. In some implementations, an imaging system may a first image in a series of images using a first exposure setting. The imaging system may detect one or more regions of interest (ROI) in the first image. The imaging system may determine a first brightness value associated with a first ROI in the first image. The imaging system may determine a second exposure setting based on the first brightness value. The imaging system may capture a second image in the series of images, subsequent to the first image, using the second exposure setting.

Abstract: Methods and systems for close detection are provided. An input device has an opened state and a closed state. The input device includes a first portion comprising a touch sensor having a sensing area. The touch sensor includes a plurality of transmitter electrodes and a plurality of receiver electrodes. The input device also has a second portion comprising a plurality of features. A sensor circuit is configured to drive the plurality of transmitter electrodes with sensing signals and receive resulting signals from the plurality of receiver electrodes. A processing system is configured to generate an acquired image from at least a portion of the sensing area; compare the acquired image to a reference image, wherein the reference image includes data corresponding to the plurality of features; and determine whether the input device is in the closed state based on the comparison of the acquired image to the reference image.

Abstract: This disclosure provides methods, devices, and systems for active noise control (ANC). The present implementations more specifically relate to filter adaptation techniques for recursively training an infinite impulse response (IIR) filter to convert a reference audio signal to an anti-noise signal. In some aspects, an ANC system may record, via a feedforward microphone, a reference audio signal representing external noise, and may further record, via a feedback microphone, an error signal representing residual noise resulting from passive attenuation of the external noise via a primary path between the microphones. The ANC system may further estimate an IIR filter that converts the reference audio signal to the error signal based on a secondary path. In some implementations, the ANC system may recursively update the IIR filter coefficients based, at least in part, on the coefficients associated with previous frames of the reference audio signal and the error signal.

Abstract: A method for performing a scanning process using adaptive scanning, comprising: driving, by a processing system of an input device, one or more of a first set of electrodes to generate first sensing signals that are detectable by a second set of electrodes; obtaining first resulting signals associated with the first sensing signals via the second set of electrodes; determining, based on the first resulting signals, a first subset of the second set of electrodes that detected an input object on a display device; driving one or more of the first subset of the second set of electrodes that detected the input object to generate second sensing signals that are detectable by the first set of electrodes; and obtaining second resulting signals associated with the second sensing signals via the first set of electrodes.

Abstract: This disclosure provides methods, devices, and systems for audio signal processing. The present implementations more specifically relate to low-latency speech enhancement. In some aspects, a speech enhancement system may receive a number (B) of frames of a signal, where each of the B frames include a number (N) of time-domain samples. The speech enhancement system may transform the B*N time-domain samples into B*N first frequency-domain samples based on an N-point fast Fourier transform (FFT), and may further transform the B*N first frequency-domain samples into B*N second frequency-domain samples based on a B-point FFT. The speech enhancement system may determine a probability of speech in the signal based at least in part on the B*N second frequency-domain samples. In some implementations, the speech enhancement system may decimate the B*N second frequency-domain samples by a factor (D), and the probability of speech is determined based on the B*N/D decimated second frequency-domain samples.

Abstract: A method and a system for operating an artificial intelligence (AI) assistant. In some implementations, a method may include receiving a user input, the user input including a user question; generating a first vector representing the user question; matching the first vector to a second vector representing a question stored in a database of questions and answers at the computing device, the database of questions and answers including a plurality of questions extracted from a knowledge base associated with the computing device and respective answers associated with the plurality of questions; and selectively escalating the user question for processing by a machine learning model.

Abstract: A sensing device and method for driving segmented electrodes is provided. The sensing device includes a sensing area having a first sensing region and a second sensing region. The sensing device also includes a plurality of receiver electrodes comprising a first set of receiver electrodes disposed in the first sensing region and a second set of receiver electrodes disposed in the second sensing region and a plurality of segmented transmitter electrodes. Each segmented transmitter electrode comprises a first transmitter electrode and a second transmitter electrode. The first transmitter electrode is isolated from the second transmitter electrode such that the first transmitter electrode is separately controllable from the second transmitter electrode. The first transmitter electrode is disposed in the first sensing region and the second transmitter electrode is disposed in the second sensing region.

Abstract: A memory device having a plurality of groups of words is provided. Each group of the plurality of groups of words are stored in a contiguous section of a memory and each group of the plurality of groups of words includes a plurality of words. The plurality of words each have a plurality of elements with a same set of index values for a first dimension of a tensor and a same set of index vales for a second dimension of a tensor. The plurality of words have a different index value for a third dimension of the tensor.

Abstract: This disclosure provides methods, devices, and systems for video coding. The present implementations more specifically relate to autoencoders that support infer-frame coding in the latent domain. A video encoder may convert a frame of video from the pixel domain to the latent domain based on a machine learning model. For example, the machine learning model may be trained to transform the video frame into a tensor of latent attributes. In some aspects, the video encoder may combine the resulting tensor with a tensor of latent attributes associated with a previously-encoded video frame and transform the resulting tensor into a vector that includes latent attributes from both the current video frame and the previous video frame based on an inter-coding transform. More specifically, the inter-coding transform may reduce a dimensionality of the combined tensor so that the resulting vector is smaller or more compressible than the original tensor of latent attributes.

Abstract: A method for performing a reconstructed temperature stable profile sensing scheme is provided. The method comprises: based on driving a first subset of a plurality of electrodes, obtaining first resulting signals from a second subset of the plurality of electrodes; based on driving both the first subset and the second subset of the plurality of electrodes, obtaining second resulting signals from the second subset of the plurality of electrodes; determining a reconstructed temperature stable profile based on a mutual capacitance sensing profile associated with the first resulting signals, an absolute capacitance sensing (ABS) profile associated with the second resulting signals, and a reconstructed temperature stable parameter; and performing object detection based on the reconstructed temperature stable profile.

Abstract: A system for classifying touch data is provided. The system comprises: a plurality of sensing elements; and a processing system. The processing system is configured to: receive touch data from a current user via the plurality of sensing elements; determine a first set of classifier parameters corresponding to a current user based on usage data, wherein the usage data comprises the touch data from the current user, and apply the first set of classifier parameters to classify subsequent touch data.