Spiking Neural Network for Ultra-low-latency and High-accurate Object Detection

TL;DR

This paper introduces a novel SNN-based object detection model that significantly reduces latency and improves accuracy by employing timesteps compression and spike-time-dependent coding, outperforming previous models on standard datasets.

Contribution

The paper presents SUHD, the deepest spike-based object detection model with ultra-low latency and high accuracy, achieved through innovative conversion and coding techniques.

Findings

Achieves 750x fewer timesteps than Spiking-YOLO on MS COCO

Improves mean average precision by 30% on MS COCO

Demonstrates state-of-the-art performance with ultra-low latency

Abstract

Spiking Neural Networks (SNNs) have garnered widespread interest for their energy efficiency and brain-inspired event-driven properties. While recent methods like Spiking-YOLO have expanded the SNNs to more challenging object detection tasks, they often suffer from high latency and low detection accuracy, making them difficult to deploy on latency sensitive mobile platforms. Furthermore, the conversion method from Artificial Neural Networks (ANNs) to SNNs is hard to maintain the complete structure of the ANNs, resulting in poor feature representation and high conversion errors. To address these challenges, we propose two methods: timesteps compression and spike-time-dependent integrated (STDI) coding. The former reduces the timesteps required in ANN-SNN conversion by compressing information, while the latter sets a time-varying threshold to expand the information holding capacity. We also present a SNN-based ultra-low latency and high accurate object detection model (SUHD) that achieves state-of-the-art performance on nontrivial datasets like PASCAL VOC and MS COCO, with about remarkable 750x fewer timesteps and 30% mean average precision (mAP) improvement, compared to the Spiking-YOLO on MS COCO datasets. To the best of our knowledge, SUHD is the deepest spike-based object detection model to date that achieves ultra low timesteps to complete the lossless conversion.