Nvidia Corporation

Anil Ubale

Parthasarathy Sriram

Fnu Ratnesh Kumar

Yan Zhai

Subhashree Radhakrishnan

Partha Sriram

Tayfun Gurel

Varun Praveen

Input layers of an element-wise operation in a neural network can be pruned such that the shape (e.g., the height, the width, and the depth) of the pruned layers matches. A pruning engine identifies all of the input layers into the element-wise operation. For each set of corresponding neurons in the input layers, the pruning engine equalizes the metrics associated with the neurons to generate an equalized metric associated with the set. The pruning engine prunes the input layers based on the equalized metrics generated for each unique set of corresponding neurons.

Pruning neural networks that include element-wise operations

In various examples, sensor data—such as masked sensor data—may be used as input to a machine learning model to determine a confidence for object to person associations. The masked sensor data may focus the machine learning model on particular regions of the image that correspond to persons, objects, or some combination thereof. In some embodiments, coordinates corresponding to persons, objects, or combinations thereof, in addition to area ratios between various regions of the image corresponding to the persons, objects, or combinations thereof, may be used to further aid the machine learning model in focusing on important regions of the image for determining the object to person associations.

Determining associations between objects and persons using machine learning models

In various examples, sensor data&#x2014;such as masked sensor data&#x2014;may be used as input to a machine learning model to determine a confidence for object to person associations. The masked sensor data may focus the machine learning model on particular regions of the image that correspond to persons, objects, or some combination thereof. In some embodiments, coordinates corresponding to persons, objects, or combinations thereof, in addition to area ratios between various regions of the image corresponding to the persons, objects, or combinations thereof, may be used to further aid the machine learning model in focusing on important regions of the image for determining the object to person associations.

A method for processing a bitstream starts by shifting a bitstream of a first sample of a signal into a buffer. The buffer also holds bits of one or more additional bitstreams for one or more additional samples of the signal. Bits of a first half of the buffer are incrementally compared to corresponding bits of a second half of the buffer. Each bit of the first half of the buffer is compared to a corresponding bit of the second half of the buffer. A computation is performed on each bit of the first half of the buffer that is equal to a corresponding bit of the second half of the buffer. The results of the computations are summed to determine an output value for the first sample of the signal.

Efficient digital microphone receiver process and system

Voice trigger. In accordance with a first method embodiment, a long term average audio energy is determined based on a one-bit pulse-density modulation bit stream. A short term average audio energy is determined based on the one-bit pulse-density modulation bit stream. The long term average audio energy is compared to the short term average audio energy. Responsive to the comparing, a voice trigger signal is generated if the short term average audio energy is greater than the long term average audio energy. Determining the long term average audio energy may be performed independent of any decimation of the bit stream.

Voice trigger

Methods and systems for transcoding input audio data in a first encoding format to generate audio data in a second encoding format, and filterbanks for use in such systems. Some such systems include a combined synthesis and analysis filterbank (configured to generate transformed frequency-band coefficients indicative of at least one sample of the input audio data by transforming frequency-band coefficients in a manner equivalent to upsampling the frequency-band coefficients and filtering the resulting up-sampled values to generate the transformed frequency-band coefficients, where the frequency-band coefficients are partially decoded versions of input audio data that are indicative of the at least one sample) and a processing subsystem configured to generate transcoded audio data in the second encoding format in response to the transformed frequency-band coefficients. Some such methods include the steps of: generating frequency-band coefficients indicative of at least one sample of input audio data by partially decoding frequency coefficients of the input audio data; generating transformed frequency-band coefficients indicative of the at least one sample of the input audio data by transforming the frequency-band coefficients in a manner equivalent to upsampling the frequency-band coefficients to generate up-sampled values and filtering the up-sampled values; and in response to the transformed frequency-band coefficients, generating the transcoded audio data so that the transcoded audio data are indicative of each sample of the input audio data.

Growing community of inventors

Cupertino, CA, United States of America

Anil Ubale