Max Robotics — Global Service Robot Directory + US Certification

About

Amaury Depierre’s research focuses on a critical challenge in robotics: enabling robots to grasp objects with human-like dexterity. His primary contributions lie in robotic grasp detection and prediction, leveraging deep neural networks to improve how robots interact with their environment. Depierre is best known for creating the **Jacquard dataset** (2018), a large-scale, open-source resource containing over 50,000 synthetic images of objects with annotated grasp locations. This dataset has become a cornerstone in the field, amassing over **360 citations** and enabling researchers to train more robust grasp-detection models without the prohibitive cost of manual labeling. Building on this foundation, his work on “Scoring Graspability based on Grasp Regression” (2021), published at the prestigious **IEEE International Conference on Robotics and Automation (ICRA)**, introduced a novel method that jointly predicts a graspability score and regresses grasp offsets. This approach significantly improved grasp prediction accuracy by optimizing the correlation between these two outputs. With additional papers refining these techniques (totaling over **400 citations**), Depierre’s research has advanced the practical deployment of robotic grasping in real-world applications, from manufacturing to service robotics, making him a key figure in modern manipulation research.