Max Robotics — Global Service Robot Directory + US Certification

About

Guillermo Garcia-Hernando is a computer vision researcher whose work spans two interconnected domains: hand action recognition and 6D object pose estimation. He is perhaps best known for spearheading the **First-Person Hand Action Benchmark**, a large-scale RGB-D dataset comprising over 100,000 frames across 45 daily hand action categories and 26 objects, which has amassed over 500 citations since its 2018 publication and become a foundational resource for researchers studying egocentric hand-object interactions. This benchmark demonstrated the power of 3D hand pose annotations in advancing action recognition from a first-person perspective. Beyond hand analysis, Garcia-Hernando has made significant contributions to object pose recovery, authoring a comprehensive review of 6D pose estimation methods — from 3D bounding box detectors to full 6D estimators — that has attracted over 100 citations. His research also extends into intelligent, autonomous perception, exploring how deep reinforcement learning can guide strategic camera movements to improve multi-object pose estimation in cluttered, real-world robotic and augmented reality scenarios. Together, his body of work reflects a strong commitment to bridging fundamental computer vision research with practical applications in robotics and human-computer interaction.