Lessons and Winning Solutions in Industrial Object Detection and Pose Estimation from the 2025 Bin-Picking Perception Challenge

Ziqin Huang, Chengxi Li, Yingyue Li, Xingyu Liu, Chenyangguang Zhang, Ruida Zhang, Bowen Fu, Xinggang Hu, Yun Qu, Mengge Liu, Yixiu Mao, Wendong Huang, Gu Wang, Xiangyang Ji; Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops, 2025, pp. 2429-2435

Abstract

This paper analyzes the challenges encountered in object pose estimation tasks within industrial environments, based on our winning solutions in the 2025 Perception Challenge for Bin-Picking. We discuss several strategies employed during the competition and highlight two unexpected observations: (1) methods trained on object-specific datasets performed worse than those trained on unseen data; and (2) evaluation results varied significantly depending on the chosen metrics. Through a detailed analysis of these findings, we aim to provide researchers with a deeper understanding of the complexities involved in industrial object pose estimation and offer insights to improve the practical deployment of such systems.

Related Material

[pdf]
[bibtex]
@InProceedings{Huang_2025_ICCV, author = {Huang, Ziqin and Li, Chengxi and Li, Yingyue and Liu, Xingyu and Zhang, Chenyangguang and Zhang, Ruida and Fu, Bowen and Hu, Xinggang and Qu, Yun and Liu, Mengge and Mao, Yixiu and Huang, Wendong and Wang, Gu and Ji, Xiangyang}, title = {Lessons and Winning Solutions in Industrial Object Detection and Pose Estimation from the 2025 Bin-Picking Perception Challenge}, booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops}, month = {October}, year = {2025}, pages = {2429-2435} }