9 papers
GORDON: Graph-based Object-centric Rewards for Decomposition of Long-Horizon Manipulation
Andrea Protopapa, Davide Buoso, Francesca Pistilli +2
Learning long-horizon manipulation skills with reinforcement learning remains challenging due to the complexity of reward design, the limited guidance of sparse rewards, and the hi…
Learning to Evolve Scenes: Reasoning about Human Activities with Scene Graphs
Francesca Pistilli, Simone Alberto Peirone, Giuseppe Averta
Understanding human behavior while interacting with the surrounding world is crucial for many applications of embodied AI. First-person videos are particularly informative for this…
HiERO-StepG @ Ego4D Step Grounding Challenge: hierarchical activity understanding enables zero-shot step grounding
Andrea Zenotto, Simone Alberto Peirone, Francesca Pistilli +1
Procedural activities follow well-defined structures: whether we consider a cooking recipe or a mechanic repairing a car, these activities naturally decompose in a hierarchy of ste…
GAP: Geometric Anchor Pre-training for Data-Efficient Visuomotor Learning of Manipulation Tasks
Davide Buoso, Andrea Protopapa, Stefano Di Carlo +2
Learning visuomotor policies from scarce expert demonstrations remains a core challenge in robotic manipulation. A primary hurdle lies in distilling high-dimensional RGB representa…
Learning reusable concepts across different egocentric video understanding tasks
Simone Alberto Peirone, Francesca Pistilli, Antonio Alliegro +2
Our comprehension of video streams depicting human activities is naturally multifaceted: in just a few moments, we can grasp what is happening, identify the relevance and interacti…
HiERO: understanding the hierarchy of human behavior enhances reasoning on egocentric videos
Simone Alberto Peirone, Francesca Pistilli, Giuseppe Averta
Human activities are particularly complex and variable, and this makes challenging for deep learning models to reason about them. However, we note that such variability does have a…