Thibault Groueix

Papers

Total Citations

H-Index

About

Thibault Groueix is a leading researcher in 3D computer vision and human modeling, with a focus on advancing temporal understanding of human pose and shape. His work bridges deep learning and geometric processing, particularly through transformer-based architectures. In his highly cited work "PoseBERT: A Generic Transformer Module for Temporal 3D Human Modeling" (2022, 18 citations), Groueix introduced a novel transformer module that addresses the critical challenge of limited annotated video data for human pose estimation. This contribution enables state-of-the-art temporal modeling without relying on costly pseudo-ground truth annotations, making 3D human pose estimation more accessible and robust. Beyond this, Groueix has made foundational contributions to learning 3D shape representations, including work on atlas-based surface reconstruction and point cloud processing. His research has been recognized for its practical impact on animation, augmented reality, and human-computer interaction. With a publication record that continues to shape how machines understand and model the human body in motion, Groueix stands as an influential voice in the intersection of transformer architectures and 3D vision.