Real Time On Sensor Gait Phase Detection with 0.5KB Deep Learning Model

TL;DR

This paper introduces a highly compact deep learning model for real-time gait phase detection on sensors, achieving high accuracy with minimal computational resources suitable for low-power microcontrollers.

Contribution

A novel segmentation-based gait detection model significantly smaller and more efficient than existing CNNs, enabling on-sensor real-time processing.

Findings

Model size is only 0.5KB.

Achieves 95.9% accuracy.

Operates with 67K operations per second.

Abstract

Gait phase detection with convolution neural network provides accurate classification but demands high computational cost, which inhibits real time low power on-sensor processing. This paper presents a segmentation based gait phase detection with a width and depth downscaled U-Net like model that only needs 0.5KB model size and 67K operations per second with 95.9% accuracy to be easily fitted into resource limited on sensor microcontroller.