EVA$^2$: Exploiting Temporal Redundancy in Live Computer Vision

TL;DR

This paper introduces EVA$^2$, a hardware-accelerated method that leverages temporal redundancy in live video to significantly reduce energy consumption in CNN-based vision tasks while maintaining high accuracy.

Contribution

It presents a novel activation motion compensation algorithm that exploits temporal redundancy, integrated into CNN accelerators for improved efficiency in real-time vision.

Findings

Reduces energy per frame by up to 87.6%.

Maintains less than 1% accuracy loss.

Uses adaptive key frame rate for efficiency-accuracy trade-off.

Abstract

Hardware support for deep convolutional neural networks (CNNs) is critical to advanced computer vision in mobile and embedded devices. Current designs, however, accelerate generic CNNs; they do not exploit the unique characteristics of real-time vision. We propose to use the temporal redundancy in natural video to avoid unnecessary computation on most frames. A new algorithm, activation motion compensation, detects changes in the visual input and incrementally updates a previously-computed output. The technique takes inspiration from video compression and applies well-known motion estimation techniques to adapt to visual changes. We use an adaptive key frame rate to control the trade-off between efficiency and vision quality as the input changes. We implement the technique in hardware as an extension to existing state-of-the-art CNN accelerator designs. The new unit reduces the average energy per frame by 54.2%, 61.7%, and 87.6% for three CNNs with less than 1% loss in vision accuracy.