Shikun (Oliver) Ban

Stoked to have presented our #ECCV2024 paper "Real-time Holistic Robot Pose Estimation with Unknown States" @eccvconf. We estimate camera-to-robot pose without known GT joint states as additional inputs. Instead, we estimate both robot pose and joint states holistically (p1).

Excited to share our #ECCV2024 work on real-time robot pose estimation using just RGB images—no joint states needed. It’s 12x faster with state-of-the-art accuracy. Catch us at poster 134 on Oct 2, 16:30 – 18:30! 🚀🦾

🍷 🐍 biiig isolation 🐍 🍷