Detecting Traffic Lights by Single Shot Detection

TL;DR

This paper introduces a deep learning single shot detection method tailored for traffic light detection, achieving high accuracy and real-time performance, especially for small objects, outperforming traditional hand-crafted feature approaches.

Contribution

It adapts the SSD framework to effectively detect very small traffic lights, enabling real-time detection with high accuracy and low false positives.

Findings

High detection accuracy on DTLD dataset

Real-time detection at 10 fps on Nvidia Titan Xp

Effective detection of objects smaller than ten pixels

Abstract

Recent improvements in object detection are driven by the success of convolutional neural networks (CNN). They are able to learn rich features outperforming hand-crafted features. So far, research in traffic light detection mainly focused on hand-crafted features, such as color, shape or brightness of the traffic light bulb. This paper presents a deep learning approach for accurate traffic light detection in adapting a single shot detection (SSD) approach. SSD performs object proposals creation and classification using a single CNN. The original SSD struggles in detecting very small objects, which is essential for traffic light detection. By our adaptations it is possible to detect objects much smaller than ten pixels without increasing the input image size. We present an extensive evaluation on the DriveU Traffic Light Dataset (DTLD). We reach both, high accuracy and low false positive rates. The trained model is real-time capable with ten frames per second on a Nvidia Titan Xp. Code has been made available at https://github.com/julimueller/tl_ssd.