Active Observation for Robust Visual Servoing in Robotic Manipulation
Zhilun Hu, Yuqi Cheng, Yuxiang Tan, Wen-pan Li, Weiming Shen.
IEEE CASE 2026 · Accepted (Paper ThCT6.3) · First Author.
Abstract & Method Overview
During terminal robotic grasping, the manipulator or gripper frequently obscures the target object just when closed-loop visual feedback is most critical. This work proposes a dual-arm active observation framework in which one manipulator actively positions an eye-in-hand RGB-D camera while the second manipulator executes the grasp. By combining a two-stage viewpoint strategy, target–gripper relative image error feedback, and occlusion-triggered viewpoint exploration, the system preserves or recovers clear co-visibility throughout physical interaction.
BibTeX
@inproceedings{hu2026activeobservation,
title = {Active Observation for Robust Visual Servoing in Robotic Manipulation},
author = {Hu, Zhilun and Cheng, Yuqi and Tan, Yuxiang and Li, Wen-pan and Shen, Weiming},
booktitle = {2026 IEEE International Conference on Automation Science and Engineering (CASE)},
year = {2026},
note = {Accepted, Paper ThCT6.3}
}