Marcel Torné Villasevil

 

Publications

Reconciling Reality through Simulation: A Real-To-Sim-to-Real Approach for Robust Manipulation.
Marcel Torne Villasevil , Anthony Simeonov, Zechu Li, April Chan, Tao Chen, Abhishek Gupta*, Pulkit Agrawal*.
Robotics: Science and Systems 2024 / Website / Paper / Video
Media Coverage: Tech Crunch / MIT CSAIL Spotlight Video

Breadcrumbs to the Goal: Goal-Conditioned Exploration from Human-in-the-loop feedback.
Marcel Torne Villasevil , Max Balsells i Pamies, Zihan Wang, Samedh Desai, Tao Chen, Pulkit Agrawal, Abhishek Gupta.
NeurIPS 2023 / Website / Paper / Code / Talk
Media Coverage: MIT News Spotlight

Autonomous Robotic Reinforcement Learning with Asynchronous Human Feedback.
Max Balsells i Pamies*, Marcel Torne Villasevil* , Zihan Wang*, Samedh Desai, Pulkit Agrawal, Abhishek Gupta.
CoRL 2023 / Website / Paper / Code
Media Coverage: TechXplore

Master Thesis: Goal Conditioned Exploration from Human in the loop Feedback
Marcel Torne Villasevil. Thesis advisor: Prof. Pulkit Agrawal
Master of Engineering Degree at Harvard University, thesis done while at MIT at the Improbable-AI Lab
Master's Thesis

(*) indicates equal contribution

Open Source Projects

DISCO - Distributed Collaborative Machine Learning
Project started by Marcel Torne Villasevil and Paul Mansat, supervised by Professor Martin Jaggi. (2021)
Code available at github.com/epfml/disco and Bachelor's thesis available here.
105 🌟 stars on GitHub

Course Projects

Design Choices for Dual-arm Robotic Manipulation Control
Project developed as part of the final project for the Robotic Manipulation (6.4212) class by Prof. Russ Tedrake at MIT, Fall 2022.
Project report available here.

Behavior Cloning for Mobile Manipulation
Project developed as part of the final project for the Computational Sensorimotor Learning (6.484) class by Prof. Pulkit Agrawal at MIT, Spring 2022.
Project report available here.

Multi-Agent Reinforcement Learning for Collective Transport
Project developed as part of the final project for the Biologically-inspired Multi-Agent Systems (CS289) class by Prof. Radhika Nagpal at Harvard University, Fall 2021.
Project report available here.