CBinfer: Change-Based Inference for Convolutional Neural Networks on Video Data

TL;DR

This paper introduces CBinfer, a change-based CNN evaluation method for static-camera video data, significantly increasing processing speed and energy efficiency while maintaining high accuracy without retraining.

Contribution

The paper presents a novel change-based inference algorithm for CNNs on video data, exploiting pixel change sparsity to improve speed and energy efficiency without retraining.

Findings

8.6x speed-up over cuDNN baseline

Less than 0.1% accuracy loss

10x higher energy efficiency

Abstract

Extracting per-frame features using convolutional neural networks for real-time processing of video data is currently mainly performed on powerful GPU-accelerated workstations and compute clusters. However, there are many applications such as smart surveillance cameras that require or would benefit from on-site processing. To this end, we propose and evaluate a novel algorithm for change-based evaluation of CNNs for video data recorded with a static camera setting, exploiting the spatio-temporal sparsity of pixel changes. We achieve an average speed-up of 8.6x over a cuDNN baseline on a realistic benchmark with a negligible accuracy loss of less than 0.1% and no retraining of the network. The resulting energy efficiency is 10x higher than that of per-frame evaluation and reaches an equivalent of 328 GOp/s/W on the Tegra X1 platform.