Fast Person Detection Using YOLOX With AI Accelerator For Train Station Safety

TL;DR

This work tackles the safety challenge of detecting passengers on train platforms to prevent unsafe crossings at the yellow line. It proposes a fast, edge-based detection system using YOLOX deployed on the Hailo-8 AI accelerator, with COCO-based evaluation and an edge-runtime deployment workflow. Experimental results show the Hailo-8 achieves higher accuracy (over 12%) and lower latency (about 20 ms) compared to Jetson Orin Nano, validating the approach for real-time safety warnings. The study demonstrates practical applicability at a metro station, highlighting the potential for low-latency, energy-efficient safety systems in rail environments and informing future edge-AI deployments.

Abstract

Recently, Image processing has advanced Faster and applied in many fields, including health, industry, and transportation. In the transportation sector, object detection is widely used to improve security, for example, in traffic security and passenger crossings at train stations. Some accidents occur in the train crossing area at the station, like passengers uncarefully when passing through the yellow line. So further security needs to be developed. Additional technology is required to reduce the number of accidents. This paper focuses on passenger detection applications at train stations using YOLOX and Edge AI Accelerator hardware. the performance of the AI accelerator will be compared with Jetson Orin Nano. The experimental results show that the Hailo-8 AI hardware accelerator has higher accuracy than Jetson Orin Nano (improvement of over 12%) and has lower latency than Jetson Orin Nano (reduced 20 ms).

Figures (8)

• Figure 1: YOLOX Structure
• Figure 2: Model Build Environment
• Figure 3: Run-Time Architecture
• Figure 4: Detection process
• Figure 5: Object Height