Ultra-Fast Adaptive Track Detection Network

TL;DR

This paper introduces an ultra-fast adaptive railway track detection network that balances high speed and accuracy, utilizing specialized branches for perspective and coordinate localization, achieving state-of-the-art performance.

Contribution

The paper presents a novel adaptive detection network with specialized branches for perspective and coordinate localization, improving speed and accuracy over existing models.

Findings

Achieved 98.68% F1 score on SRail dataset

Reaches detection speeds up to 473 FPS

Outperforms state-of-the-art in both speed and accuracy

Abstract

Railway detection is critical for the automation of railway systems. Existing models often prioritize either speed or accuracy, but achieving both remains a challenge. To address the limitations of presetting anchor groups that struggle with varying track proportions from different camera angles, an ultra-fast adaptive track detection network is proposed in this paper. This network comprises a backbone network and two specialized branches (Horizontal Coordinate Locator and Perspective Identifier). The Perspective Identifier selects the suitable anchor group from preset anchor groups, thereby determining the row coordinates of the railway track. Subsequently, the Horizontal Coordinate Locator provides row classification results based on multiple preset anchor groups. Then, utilizing the results from the Perspective Identifier, it generates the column coordinates of the railway track. This network is evaluated on multiple datasets, with the lightweight version achieving an F1 score of 98.68% on the SRail dataset and a detection rate of up to 473 FPS. Compared to the SOTA, the proposed model is competitive in both speed and accuracy. The dataset and code are available at https://github.com/idnihai/UFATD