Amazon Technologies, Inc.

Shiva Kumar Sundaram

Arindam Mandal

Minhua Wu

Kenichi Kumatani

Spyridon Matsoukas

Ramya Gopalan

Bjorn Hoffmeister

Chao Wang

Shiv Naga Prasad Vitaladevuni

Nikko Strom

Anirudh Raju

Philip Ryan Hilmes

Amit Singh Chhetri

Michael Patrick Meaney

Sankaran Panchapagesan

Rui Wang

Tae Jin Park

Systems and methods for utilizing microphone array information for acoustic modeling are disclosed. Audio data may be received from a device having a microphone array configuration. Microphone configuration data may also be received that indicates the configuration of the microphone array. The microphone configuration data may be utilized as an input vector to an acoustic model, along with the audio data, to generate phoneme data. Additionally, the microphone configuration data may be utilized to train and/or generate acoustic models, select an acoustic model to perform speech recognition with, and/or to improve trigger sound detection.

Speech processing optimizations based on microphone array

A speech-capture device can capture audio data during wakeword monitoring and use the audio data to determine if a user is present nearby the device, even if no wakeword is spoken. Audio such as speech, human originating sounds (e.g., coughing, sneezing), or other human related noises (e.g., footsteps, doors closing) can be used to detect audio. Audio frames are individually scored as to whether a human presence is detected in the particular audio frames. The scores are then smoothed relative to nearby frames to create a decision for a particular frame. Presence information can then be sent according to a periodic schedule to a remote device to create a presence 'heartbeat' that regularly identifies whether a user is detected proximate to a speech-capture device.

User presence detection

Techniques for speech processing using a deep neural network (DNN) based acoustic model front-end are described. A new modeling approach directly models multi-channel audio data received from a microphone array using a first model (e.g., multi-geometry/multi-channel DNN) that is trained using a plurality of microphone array geometries. Thus, the first model may receive a variable number of microphone channels, generate multiple outputs using multiple microphone array geometries, and select the best output as a first feature vector that may be used similarly to beamformed features generated by an acoustic beamformer. A second model (e.g., feature extraction DNN) processes the first feature vector and transforms it to a second feature vector having a lower dimensional representation. A third model (e.g., classification DNN) processes the second feature vector to perform acoustic unit classification and generate text data. The DNN front-end enables improved performance despite a reduction in microphones.

Deep multi-channel acoustic modeling using multiple microphone array geometries

Techniques for speech processing using a deep neural network (DNN) based acoustic model front-end are described. A new modeling approach directly models multi-channel audio data received from a microphone array using a first model (e.g., multi-geometry/multi-channel DNN) that includes a frequency aligned network (FAN) architecture. Thus, the first model may perform spatial filtering to generate a first feature vector by processing individual frequency bins separately, such that multiple frequency bins are not combined. The first feature vector may be used similarly to beamformed features generated by an acoustic beamformer. A second model (e.g., feature extraction DNN) processes the first feature vector and transforms it to a second feature vector having a lower dimensional representation. A third model (e.g., classification DNN) processes the second feature vector to perform acoustic unit classification and generate text data. The DNN front-end enables improved performance despite a reduction in microphones.

Deep multi-channel acoustic modeling using frequency aligned network

Techniques for speech processing using a deep neural network (DNN) based acoustic model front-end are described. A new modeling approach directly models multi-channel audio data received from a microphone array using a first model (e.g., multi-channel DNN) that takes in raw signals and produces a first feature vector that may be used similarly to beamformed features generated by an acoustic beamformer. A second model (e.g., feature extraction DNN) processes the first feature vector and transforms it to a second feature vector having a lower dimensional representation. A third model (e.g., classification DNN) processes the second feature vector to perform acoustic unit classification and generate text data. These three models may be jointly optimized for speech processing (as opposed to individually optimized for signal enhancement), enabling improved performance despite a reduction in microphones and a reduction in bandwidth consumption during real-time processing.

Deep multi-channel acoustic modeling

A system capable of detecting human presence based on output from a model-free detector and model-based detector(s). For example, the model-free detector may identify acoustic events and the model-based detectors can determine specific types of acoustic events and whether the acoustic events are associated with human activity. Using output from the model-based detectors, a device may confirm that an acoustic event identified by the model-free detector is associated with human activity or may determine that the acoustic event is associated with non-human activity and can be ignored. Thus, the device may detect human presence based on a wide variety of noises while reducing a number of false positives associated with the model-free detector.

Hybrid audio-based presence detection

A neural network acoustic model is trained to be robust and produce accurate output when used to process speech signals having acoustic interference. The neural network acoustic model can be trained using a source-separation process by which, in addition to producing the main acoustic model output for a given input, the neural network generates predictions of the separate speech and interference portions of the input. The parameters of the neural network can be adjusted to jointly optimize all three outputs (e.g., the main acoustic model output, the speech signal prediction, and the interference signal prediction), rather than only optimizing the main acoustic model output. Once trained, output layers for the speech and interference signal predictions can be removed from the neural network or otherwise disabled.

Robust neural network acoustic model with side task prediction of reference signals

A neural network model, such as a deep neural network (DNN), is trained using many speech examples to perform beam selection in a microphone array-based speech processing system. The DNN is trained using many different speech examples that are labeled with position or direction information relative to a training microphone array. The DNN may then be trained to recognize a direction of incoming speech so that at runtime the trained DNN may process input audio data from a microphone array and may output to a beam selector an indicator of the desired beam that may be selected for further processing. The DNN may be configured to output a beam index and/or coordinates (or other position data) corresponding to an estimated location of the detected speech. The DNN may also be configured to output acoustic unit data corresponding to speech units (for example corresponding to phonemes, senons, etc. such as those of a detected wakeword or other word).

Neural network based beam selection

Architectures and techniques for selecting a voice-enabled device to handle audio input that is detected by multiple voice-enabled devices are described herein. In some instances, multiple voice-enabled devices may detect audio input from a user at substantially the same time, due to the voice-enabled devices being located within proximity to the user. The architectures and techniques may analyze a variety of audio signal metric values for the voice-enabled devices to designate a voice-enabled device to handle the audio input.

Growing community of inventors

Mountain View, CA, United States of America

Shiva Kumar Sundaram