Active Perception Improved Grasping In

Centre PhD Researcher Doug Morrison tests and validates his world-first

Active Perception

Paper: https://arxiv.org/abs/2106.01866 IRL-Lab: https://www.ai.rug.nl/irl-lab/ To aid humans in everyday tasks, robots need to ...

Object

Title: Reinforcement Learning-based

In this paper, we propose a multi-view deep learning approach to handle robust object

Some objects cannot be discriminated by looking at a single view. In this video we demonstrate an

This video shows that our dexterous

CAPerMoMa: Coupled

M. Corsaro, S. Tellex, and G.D. Konidaris. Learning to Detect Multi-Modal

Demonstration of line inspection task in the Belief space. An illustrative video about the synthesis you can find here: ...

This video demonstrates the results of our approach of dynamically

A supplementary video of our paper accepted at IROS 2020: "Distributed Reinforcement Learning of Targeted

UC Berkeley AUTOLAB http://bit.ly/AUTOLAB Dex-Net 2.0: Deep Learning to Plan Robust

Paper: https://arxiv.org/abs/2104.07267 Paper website: https://www.pgrady.net/contactopt/ Abstract: Physical contact between ...

Simulation Setup ...

Vision-based Aided-

Control of Memory,

Description.

Multi-View Picking: Next-best-view Reaching for