EMDFNet: Efficient Multi-scale and Diverse Feature Network for Traffic Sign Detection

Pengyu Li,Chenhe Liu,Tengfei Li,Xinyu Wang,Shihui Zhang,Dongyang Yu

2024-08-26

Abstract:The detection of small objects, particularly traffic signs, is a critical subtask within object detection and autonomous driving. Despite the notable achievements in previous research, two primary challenges persist. Firstly, the main issue is the singleness of feature extraction. Secondly, the detection process fails to effectively integrate with objects of varying sizes or scales. These issues are also prevalent in generic object detection. Motivated by these challenges, in this paper, we propose a novel object detection network named Efficient Multi-scale and Diverse Feature Network (EMDFNet) for traffic sign detection that integrates an Augmented Shortcut Module and an Efficient Hybrid Encoder to address the aforementioned issues simultaneously. Specifically, the Augmented Shortcut Module utilizes multiple branches to integrate various spatial semantic information and channel semantic information, thereby enhancing feature diversity. The Efficient Hybrid Encoder utilizes global feature fusion and local feature interaction based on various features to generate distinctive classification features by integrating feature information in an adaptable manner. Extensive experiments on the Tsinghua-Tencent 100K (TT100K) benchmark and the German Traffic Sign Detection Benchmark (GTSDB) demonstrate that our EMDFNet outperforms other state-of-the-art detectors in performance while retaining the real-time processing capabilities of single-stage models. This substantiates the effectiveness of EMDFNet in detecting small traffic signs.

Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

### Problems the paper attempts to solve

This paper aims to address two main challenges in traffic sign detection:

1. **Singleness in feature extraction**: Traditional traffic sign detection methods are too single in feature extraction, resulting in poor detection performance for small targets in complex backgrounds.
2. **Insufficiency in multi - scale fusion**: Existing detection methods are unable to effectively perform multi - scale feature fusion when dealing with objects of different sizes or scales, which affects the detection performance.

To meet these challenges, the author proposes a new object detection network - the Efficient Multi - scale and Diverse Feature Network (EMDFNet). EMDFNet solves the problems of feature singleness and insufficient multi - scale fusion respectively by introducing the Augmented Shortcut Module (ASM) and the Efficient Hybrid Encoder (EHE). Specifically:

- **Augmented Shortcut Module (ASM)**: It enhances feature diversity by integrating different spatial - semantic information and channel - semantic information through multiple branches.
- **Efficient Hybrid Encoder (EHE)**: It combines global feature fusion and local feature interaction to generate discriminative classification features and realizes the adaptive fusion of multi - scale features.

Through a large number of experiments on the Tsinghua - Tencent 100K (TT100K) and German Traffic Sign Detection Benchmark (GTSDB) datasets, it is verified that EMDFNet significantly improves the detection performance of small traffic signs while maintaining real - time processing capabilities.