YOLO-LLTS: Real-Time Low-Light Traffic Sign Detection via Prior-Guided Enhancement and Multibranch Feature Interaction

TL;DR

This work targets robust traffic sign detection in challenging low-light conditions by presenting YOLO-LLTS, an end-to-end detector augmented with three novel modules: HRFM-SOD for high-resolution small-object feature preservation, MFIA for multi-receptive-field feature interaction, and PGFE for prior-guided feature enhancement. A new nighttime dataset CNTSSS is introduced to address data scarcity in low-light scenarios, and comprehensive experiments across TT100K-night, CNTSSS, CCTSDB2021, and GTSDB-night demonstrate state-of-the-art performance, including improvements of up to $2.7\%$ in mAP50 and $1.6\%$ in mAP50:95 on TT100K-night. Ablation studies confirm the distinct value of HRFM-SOD, MFIA, and PGFE, with combined usage yielding the largest gains while remaining suitable for edge devices. The method’s practicality is further validated through edge-device deployments showing real-time inference and robust nighttime detection, underscoring potential benefits for ADAS and autonomous driving safety. $}$

Abstract

Traffic sign detection is essential for autonomous driving and Advanced Driver Assistance Systems (ADAS). However, existing methods struggle to address the challenges of poor image quality and insufficient information under low-light conditions, leading to a decline in detection accuracy and affecting driving safety. To address this issue, we propose YOLO-LLTS, an end-to-end real-time traffic sign detection algorithm specifically designed for low-light environments. YOLO-LLTS introduces three main contributions: the HRFM-SOD module retains more information about distant or tiny traffic signs compared to traditional methods; the MFIA module interacts features with different receptive fields to improve information utilization; the PGFE module enhances detection accuracy by improving brightness, edges, contrast, and supplementing detail information. Additionally, we construct a new dataset, the Chinese Nighttime Traffic Sign Sample Set (CNTSSS), covering diverse nighttime scenarios. Experiments show that YOLO-LLTS achieves state-of-the-art performance, outperforming previous best methods by 2.7% mAP50 and 1.6% mAP50:95 on TT100K-night, 1.3% mAP50 and 1.9% mAP50:95 on CNTSSS, 7.5% mAP50 and 9.8% mAP50:95 on GTSDB-night, and superior results on CCTSDB2021. Deployment on edge devices confirms its real-time applicability and effectiveness. The code and the dataset are available at https://github.com/linzy88/YOLO-LLTS.

Figures (20)

• Figure 1: The comparison of traffic signs in normal and low-light environments. The edges of traffic signs under low light are unclear and detecting the targets becomes challenging.
• Figure 2: Application Scenarios of Traffic Sign Detection in advanced driver-assistance systems ico52112iconfinder. Vehicles equipped with ADAS systems capture real-time image information through cameras. By detecting nighttime traffic signs using YOLO-LLTS, the system alerts the driver to take corrective actions or directly controls the vehicle to enhance driving safety.
• Figure 3: Comprehensive collection of nighttime traffic sign images captured across 17 cities in China, including Beijing, Shanghai, Guangzhou, Shenzhen, Jiangmen, Chongqing, Chengdu, Nanchong, Wuhan, Changsha, Tianjin, Nanjing, Zhenjiang, Shangqiu, Shangrao, Guilin, and Jingdezhen.
• Figure 4: The CNTSSS dataset includes three categories of traffic signs: prohibitory signs, mandatory signs, and warning signs. Specifically, prohibitory signs account for 64% of the collection, mandatory signs account for 22%, and warning signs account for 14%.
• Figure 5: The distribution of object anchor box sizes in CNTSSS dataset.