Small, but important: Traffic light proposals for detecting small traffic lights and beyond

TL;DR

This paper introduces a novel traffic light detection system that significantly improves detection of small and tiny traffic lights by addressing CNN downsampling issues through multi-scale features, attention, and a new detection head.

Contribution

It presents a new detection system with a proposal generator and detection head that enhance small traffic light detection, outperforming existing methods.

Findings

At least 12.6% improvement on small and tiny traffic lights

Strong overall performance across all traffic light sizes

Evaluated on three challenging public datasets

Abstract

Traffic light detection is a challenging problem in the context of self-driving cars and driver assistance systems. While most existing systems produce good results on large traffic lights, detecting small and tiny ones is often overlooked. A key problem here is the inherent downsampling in CNNs, leading to low-resolution features for detection. To mitigate this problem, we propose a new traffic light detection system, comprising a novel traffic light proposal generator that utilizes findings from general object proposal generation, fine-grained multi-scale features, and attention for efficient processing. Moreover, we design a new detection head for classifying and refining our proposals. We evaluate our system on three challenging, publicly available datasets and compare it against six methods. The results show substantial improvements of at least $12.6\%$ on small and tiny traffic lights, as well as strong results across all sizes of traffic lights.