EgoPAT3Dv2: Predicting 3D Action Target from 2D Egocentric Vision for Human-Robot Interaction
Published in IEEE International Conference on Robotics and Automation (ICRA), 2024
Predicting where a human hand is heading is a key capability for proactive human-robot interaction. EgoPAT3Dv2 predicts the 3D action target directly from 2D egocentric video, removing the need for depth sensing at inference time and generalizing to unseen scenes and users.
Recommended citation: Fang, I., Chen, Y., Wang, Y., Zhang, J., Zhang, Q., Xu, J., He, X., Gao, W., Su, H., Li, Y., & Feng, C. (2024). EgoPAT3Dv2: Predicting 3D Action Target from 2D Egocentric Vision for Human-Robot Interaction. IEEE International Conference on Robotics and Automation (ICRA).
Download Paper