DualCam: A Novel Benchmark Dataset for Fine-grained Real-time Traffic Light Detection

TL;DR

This paper introduces DualCam, a new benchmark dataset with synchronized narrow- and wide-angle camera images for improved traffic light detection in urban environments, enhancing algorithm development for self-driving cars.

Contribution

The paper presents a novel synchronized dual-camera dataset and a post-processing algorithm that balances speed and accuracy for traffic light detection.

Findings

The dataset includes 1032 training and 813 testing image pairs.

The proposed method improves detection accuracy while maintaining real-time performance.

Synchronized video pairs enable qualitative analysis of traffic light detection.

Abstract

Traffic light detection is essential for self-driving cars to navigate safely in urban areas. Publicly available traffic light datasets are inadequate for the development of algorithms for detecting distant traffic lights that provide important navigation information. We introduce a novel benchmark traffic light dataset captured using a synchronized pair of narrow-angle and wide-angle cameras covering urban and semi-urban roads. We provide 1032 images for training and 813 synchronized image pairs for testing. Additionally, we provide synchronized video pairs for qualitative analysis. The dataset includes images of resolution 1920$\times$1080 covering 10 different classes. Furthermore, we propose a post-processing algorithm for combining outputs from the two cameras. Results show that our technique can strike a balance between speed and accuracy, compared to the conventional approach of using a single camera frame.