Enhancing Traffic Object Detection in Variable Illumination with RGB-Event Fusion

TL;DR

This paper introduces a novel fusion network leveraging RGB and event camera data to improve traffic object detection under variable illumination, achieving significant performance gains over traditional frame-based methods.

Contribution

The paper proposes SFNet, a structure-aware fusion network with RSGNet and AFCM modules, and introduces the DSEC-Det dataset for enhanced event-based traffic object detection.

Findings

Outperforms frame-based methods by 8.0% in mAP50

Generates illumination-robust representations

Provides a new dataset with 63,931 images and 208,000 labels

Abstract

Traffic object detection under variable illumination is challenging due to the information loss caused by the limited dynamic range of conventional frame-based cameras. To address this issue, we introduce bio-inspired event cameras and propose a novel Structure-aware Fusion Network (SFNet) that extracts sharp and complete object structures from the event stream to compensate for the lost information in images through cross-modality fusion, enabling the network to obtain illumination-robust representations for traffic object detection. Specifically, to mitigate the sparsity or blurriness issues arising from diverse motion states of traffic objects in fixed-interval event sampling methods, we propose the Reliable Structure Generation Network (RSGNet) to generate Speed Invariant Frames (SIF), ensuring the integrity and sharpness of object structures. Next, we design a novel Adaptive Feature Complement Module (AFCM) which guides the adaptive fusion of two modality features to compensate for the information loss in the images by perceiving the global lightness distribution of the images, thereby generating illumination-robust representations. Finally, considering the lack of large-scale and high-quality annotations in the existing event-based object detection datasets, we build a DSEC-Det dataset, which consists of 53 sequences with 63,931 images and more than 208,000 labels for 8 classes. Extensive experimental results demonstrate that our proposed SFNet can overcome the perceptual boundaries of conventional cameras and outperform the frame-based method by 8.0% in mAP50 and 5.9% in mAP50:95. Our code and dataset will be available at https://github.com/YN-Yang/SFNet.