BROWSE
Detail
Conference Paper
실감형 콘텐츠를 위한 Edge Device 환경에서의 Gated OKS 기반 실시간 Tracking 알고리즘
Cited - time in 
Share 
Share
- Issue Date
- 2026-06
- Citation
- 대한전자공학회 학술 대회 (하계) 2026, pp.1-5
- Publisher
- 대한전자공학회
- Language
- Korean
- Type
- Conference Paper
- Abstract
- In XR and immersive content environments, technologies are required to recognize and stably track users’ movements and posture changes in real time. In this paper, we propose a Gated OKS tracking algorithm that utilizes skeleton information extracted by trt_pose in a Jetson Orin NX-based edge device environment. On the Jetson Orin NX, the trt_pose-Gated OKS method achieved an average of 52.89 FPS on the MOT17 dataset, and recorded 60.32 FPS on the MOT17-06 sequence, showing the highest processing speed among the compared methods. In addition, qualitative evaluation confirmed that stable skeleton tracking is possible even under human movement, posture changes, and occlusion situations. These results demonstrate that the proposed method can achieve both real-time performance and tracking stability in edge device environments.
- KSP Keywords
- Edge devices, Human Movement, Posture changes, Processing speed, Qualitative Evaluation, Real-time, Tracking algorithm, Tracking stability, skeleton tracking