Distributed system-on-a-chip for human activity recognition

Inventors

Gu, Jie • Wei, Yijie

Assignees

Northwestern University

Abstract

In certain aspects, a system-on-a-chip (SoC) for human activity recognition includes a plurality of integrated low-noise amplifiers configured to sense electromyogram (EMG) signals. The SoC includes a mixed-signal circuitry configured to receive the EMG signals from the plurality of integrated low-noise amplifiers, wherein the mixed-signal circuitry is configured to digitalize and extract time-domain features from the EMG signals. The SoC includes an artificial intelligence (AI) core comprising a reconfigurable neural network (NN) configured to receive, from the mixed-signal circuitry, the time-domain features that were extracted, wherein the reconfigurable NN is configured to recognize a local gesture based on time-domain features that is extracted. The SoC includes an analog data path circuitry configured to sense distance measurements and to transmit the distance measurements and the local gesture that is recognized.

Core Innovation

The invention provides a system-on-a-chip for human activity recognition that senses electromyogram (EMG) signals using a plurality of integrated low-noise amplifiers. Mixed-signal circuitry receives the EMG signals from the integrated low-noise amplifiers to digitalize and extract time-domain features, and the extracted time-domain features are supplied to an artificial intelligence (AI) core comprising a reconfigurable neural network (NN).

The reconfigurable neural network is configured to recognize a local gesture based on the extracted time-domain features. An analog data path circuitry is configured to sense distance measurements and to transmit the distance measurements and the local gesture that is recognized, and the analog data path circuitry comprises an IR LED transceiver configured to transmit both the distance measurements and the local gesture that is recognized.

A distributed implementation is disclosed in which a plurality of distributed system-on-a-chips operates together for human activity recognition. Each distributed system-on-a-chip performs EMG sensing with integrated low-noise amplifiers, mixed-signal digitization and time-domain feature extraction, and local gesture recognition using a reconfigurable neural network, and the analog data path circuitry transmits the distance measurements and the local gesture that is recognized between system-on-a-chips using an IR LED transceiver.

Claims Coverage

The document includes three independent claims. The shared inventive coverage is integrated EMG sensing and mixed-signal extraction of time-domain features, a reconfigurable neural network that recognizes a local gesture from those features, and an analog data path that senses distance measurements and transmits both distance measurements and the recognized local gesture using an IR LED transceiver, including distribution between adjacent system-on-a-chips in the multi-chip system.

The inventive features are the integrated EMG sensing, mixed-signal time-domain feature extraction, reconfigurable neural network local gesture recognition, and analog distance measurement sensing with IR LED transmission of the distance measurements and recognized local gesture.

Stated Advantages

Documented Applications