publications

2026

1. arxiv preprint

   

   Milo, a Fully Autonomous Indoor/Outdoor Robotic Guide Dog

   Florian Golemo, Joanna Wolski, Joel Ruben Antony Moniz, and Christopher Pal

   arXiv preprint arXiv:2607.19530, 2026

   Abs arXiv Bib PDF Website

   Real guide dogs are incredibly expensive (50,000 USD) and take long to train. With Milo, we’ve developed an autonomous (self-contained, no cloud compute) and low-cost (2,000 USD) robot guide dog that can follow paths and avoid obstacles and pedestrians indoors and outdoors. And in order to foster open access for the blind/low-vision community, we’re open-sourcing our research.

   @article{golemo2026milo,
     title = {Milo, a Fully Autonomous Indoor/Outdoor Robotic Guide Dog},
     author = {Golemo, Florian and Wolski, Joanna and Moniz, Joel Ruben Antony and Pal, Christopher},
     journal = {arXiv preprint arXiv:2607.19530},
     url = {https://arxiv.org/abs/2607.19530},
     year = {2026},
   }

2. journal

   

   Generative Floor Plan Design with LLMs via Reinforcement Learning with Verifiable Rewards

   Luis Lara, Aristides Milios, ZhiHao Luo, Aditya Sharma, Ge Ya Luo, Christopher Beckham, Florian Golemo, and Christopher Pal

   In Findings of the Association for Computational Linguistics: ACL 2026, Jul 2026

   Abs DOI arXiv Bib

   An AI system for professional floor plan design must precisely control room dimensions and areas while respecting the desired connectivity between rooms and maintaining functional and aesthetic quality.Existing generative approaches focus primarily on respecting the requested connectivity between rooms, but do not support generating floor plans that respect numerical constraints. We introduce a text-based floor plan generation approach that fine-tunes a large language model (LLM) on real plans and then applies reinforcement learning with verifiable rewards (RLVR) to improve adherence to topological and numerical constraints while discouraging invalid or overlapping outputs.Furthermore, we design a set of constraint adherence metrics to systematically measure how generated floor plans align with user-defined constraints.Our model generates floor plans that satisfy user-defined connectivity and numerical constraints and outperforms existing methods on Realism, Compatibility, and Diversity metrics. Across all tasks, our approach achieves at least a 94% relative reduction in Compatibility compared with existing methods.Our results demonstrate that LLMs can effectively handle constraints in this setting, suggesting broader applications for text-based generative modeling.

   @inproceedings{lara-etal-2026-generative,
     title = {Generative Floor Plan Design with {LLM}s via Reinforcement Learning with Verifiable Rewards},
     author = {Lara, Luis and Milios, Aristides and Luo, ZhiHao and Sharma, Aditya and Luo, Ge Ya and Beckham, Christopher and Golemo, Florian and Pal, Christopher},
     editor = {Liakata, Maria and Moreira, Viviane P. and Zhang, Jiajun and Jurgens, David},
     booktitle = {Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026},
     month = jul,
     year = {2026},
     address = {San Diego, California, United States},
     publisher = {Association for Computational Linguistics},
     url = {https://aclanthology.org/2026.findings-acl.1326/},
     doi = {10.18653/v1/2026.findings-acl.1326},
     pages = {26612--26627},
     isbn = {979-8-89176-395-1}
   }

2025

1. patent

   Cross-modality curiosity for sparse-reward tasks

   Sai Rajeswar Mudumba, Cyril Ibrahim, Nitin Surya, Florian Golemo, David Vasquez, and Pedro O Pinheiro

   Mar 2025

   US Patent 12,254,063

   Bib

   @misc{mudumba2025cross,
     title = {Cross-modality curiosity for sparse-reward tasks},
     author = {Mudumba, Sai Rajeswar and Ibrahim, Cyril and Surya, Nitin and Golemo, Florian and Vasquez, David and Pinheiro, Pedro O},
     year = {2025},
     month = mar,
     note = {US Patent 12,254,063}
   }

2. arxiv preprint

   

   Ctrl-Crash: Controllable Diffusion for Realistic Car Crashes

   Anthony Gosselin, Ge Ya Luo, Luis Lara, Florian Golemo, Derek Nowrouzezahrai, Liam Paull, Alexia Jolicoeur-Martineau, and Christopher Pal

   arXiv preprint arXiv:2506.00227, 2025

   arXiv Bib Website

   @article{gosselin2025ctrl,
     title = {Ctrl-Crash: Controllable Diffusion for Realistic Car Crashes},
     author = {Gosselin, Anthony and Luo, Ge Ya and Lara, Luis and Golemo, Florian and Nowrouzezahrai, Derek and Paull, Liam and Jolicoeur-Martineau, Alexia and Pal, Christopher},
     journal = {arXiv preprint arXiv:2506.00227},
     year = {2025}
   }

2024

1. journal

   

   Growspace: a reinforcement learning environment for plant architecture

   Yasmeen Hitti, Ionelia Buzatu, Manuel Del Verme, Mark Lefsrud, Florian Golemo, and Audrey Durand

   Computers and Electronics in Agriculture, 2024

   Bib

   @article{hitti2024growspace,
     title = {Growspace: a reinforcement learning environment for plant architecture},
     author = {Hitti, Yasmeen and Buzatu, Ionelia and Del Verme, Manuel and Lefsrud, Mark and Golemo, Florian and Durand, Audrey},
     journal = {Computers and Electronics in Agriculture},
     volume = {217},
     pages = {108613},
     year = {2024},
     publisher = {Elsevier}
   }

2. corl

   

   Ctrl-sim: Reactive and controllable driving agents with offline reinforcement learning

   Luke Rowe, Roger Girgis, Anthony Gosselin, Bruno Carrez, Florian Golemo, Felix Heide, Liam Paull, and Christopher Pal

   arXiv preprint arXiv:2403.19918, 2024

   arXiv Bib

   @article{rowe2024ctrl,
     title = {Ctrl-sim: Reactive and controllable driving agents with offline reinforcement learning},
     author = {Rowe, Luke and Girgis, Roger and Gosselin, Anthony and Carrez, Bruno and Golemo, Florian and Heide, Felix and Paull, Liam and Pal, Christopher},
     journal = {arXiv preprint arXiv:2403.19918},
     year = {2024}
   }

3. neurips

   Dstruct2design: Data and benchmarks for data structure driven generative floor plan design

   Zhi Hao Luo, Luis Lara, Ge Ya Luo, Florian Golemo, Christopher Beckham, and Christopher Pal

   arXiv preprint arXiv:2407.15723, 2024

   arXiv Bib

   @article{luo2024dstruct2design,
     title = {Dstruct2design: Data and benchmarks for data structure driven generative floor plan design},
     author = {Luo, Zhi Hao and Lara, Luis and Luo, Ge Ya and Golemo, Florian and Beckham, Christopher and Pal, Christopher},
     journal = {arXiv preprint arXiv:2407.15723},
     year = {2024}
   }

2023

1. journal

   Visual question answering from another perspective: CLEVR mental rotation tests

   Christopher Beckham, Martin Weiss, Florian Golemo, Sina Honari, Derek Nowrouzezahrai, and Christopher Pal

   Pattern Recognition, 2023

   Bib

   @article{beckham2023visual,
     title = {Visual question answering from another perspective: CLEVR mental rotation tests},
     author = {Beckham, Christopher and Weiss, Martin and Golemo, Florian and Honari, Sina and Nowrouzezahrai, Derek and Pal, Christopher},
     journal = {Pattern Recognition},
     volume = {136},
     pages = {109209},
     year = {2023},
     publisher = {Pergamon}
   }

2. journal

   Towards learning to imitate from a single video demonstration

   Glen Berseth, Florian Golemo, and Christopher Pal

   Journal of Machine Learning Research, 2023

   Bib

   @article{berseth2023towards,
     title = {Towards learning to imitate from a single video demonstration},
     author = {Berseth, Glen and Golemo, Florian and Pal, Christopher},
     journal = {Journal of Machine Learning Research},
     volume = {24},
     number = {78},
     pages = {1--26},
     year = {2023}
   }

2022

1. corl

   Haptics-based Curiosity for Sparse-reward Tasks

   Sai Rajeswar, Cyril Ibrahim, Nitin Surya, Florian Golemo, David Vazquez, Aaron Courville, and Pedro O. Pinheiro

   In Proceedings of the 5th Conference on Robot Learning, 08–11 nov 2022

   Abs arXiv Bib PDF

   Robots in many real-world settings have access to force/torque sensors in their gripper and tactile sensing is often necessary for tasks that involve contact-rich motion. In this work, we leverage surprise from mismatches in haptics feedback to guide exploration in hard sparse-reward reinforcement learning tasks. Our approach, Haptics-based Curiosity (HaC), learns what visible objects interactions are supposed to “feel" like. We encourage exploration by rewarding interactions where the expectation and the experience do not match. We test our approach on a range of haptics-intensive robot arm tasks (e.g. pushing objects, opening doors), which we also release as part of this work. Across multiple experiments in a simulated setting, we demonstrate that our method is able to learn these difficult tasks through sparse reward and curiosity alone. We compare our cross-modal approach to single-modality (haptics- or vision-only) approaches as well as other curiosity-based methods and find that our method performs better and is more sample-efficient.

   @inproceedings{pmlr-v164-rajeswar22a,
     title = {Haptics-based Curiosity for Sparse-reward Tasks},
     author = {Rajeswar, Sai and Ibrahim, Cyril and Surya, Nitin and Golemo, Florian and Vazquez, David and Courville, Aaron and Pinheiro, Pedro O.},
     booktitle = {Proceedings of the 5th Conference on Robot Learning},
     pages = {395--405},
     year = {2022},
     editor = {Faust, Aleksandra and Hsu, David and Neumann, Gerhard},
     volume = {164},
     series = {Proceedings of Machine Learning Research},
     month = {08--11 Nov},
     publisher = {PMLR},
     url = {https://proceedings.mlr.press/v164/rajeswar22a.html}
   }

2. corl

   Haptics-based curiosity for sparse-reward tasks

   Sai Rajeswar, Cyril Ibrahim, Nitin Surya, Florian Golemo, David Vazquez, Aaron Courville, and Pedro O Pinheiro

   In Conference on Robot Learning, 2022

   Bib

   @inproceedings{rajeswar2022haptics,
     title = {Haptics-based curiosity for sparse-reward tasks},
     author = {Rajeswar, Sai and Ibrahim, Cyril and Surya, Nitin and Golemo, Florian and Vazquez, David and Courville, Aaron and Pinheiro, Pedro O},
     booktitle = {Conference on Robot Learning},
     pages = {395--405},
     year = {2022},
     organization = {PMLR}
   }

3. cvpr

   Kubric: A scalable dataset generator

   Klaus Greff, Francois Belletti, Lucas Beyer, Carl Doersch, Yilun Du, Daniel Duckworth, David J Fleet, Dan Gnanapragasam, Florian Golemo, Charles Herrmann, and others

   In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022

   Bib

   @inproceedings{greff2022kubric,
     title = {Kubric: A scalable dataset generator},
     author = {Greff, Klaus and Belletti, Francois and Beyer, Lucas and Doersch, Carl and Du, Yilun and Duckworth, Daniel and Fleet, David J and Gnanapragasam, Dan and Golemo, Florian and Herrmann, Charles and others},
     booktitle = {Proceedings of the IEEE/CVF conference on computer vision and pattern recognition},
     pages = {3749--3761},
     year = {2022}
   }

4. The sandbox environment for generalizable agent research (segar)

   R Devon Hjelm, Bogdan Mazoure, Florian Golemo, Samira Ebrahimi Kahou, Pedro Braga, Felipe Frujeri, Mihai Jalobeanu, and Andrey Kolobov

   arXiv preprint arXiv:2203.10351, 2022

   arXiv Bib

   @article{hjelm2022sandbox,
     title = {The sandbox environment for generalizable agent research (segar)},
     author = {Hjelm, R Devon and Mazoure, Bogdan and Golemo, Florian and Kahou, Samira Ebrahimi and Braga, Pedro and Frujeri, Felipe and Jalobeanu, Mihai and Kolobov, Andrey},
     journal = {arXiv preprint arXiv:2203.10351},
     year = {2022}
   }

5. Visual Question Answering From Another Perspective: CLEVR Mental Rotation Tests

   Christopher Beckham, Martin Weiss, Florian Golemo, Sina Honari, Derek Nowrouzezahrai, and Christopher Pal

   arXiv preprint arXiv:2212.01639, 2022

   arXiv Bib

   @article{beckham2022visual,
     title = {Visual Question Answering From Another Perspective: CLEVR Mental Rotation Tests},
     author = {Beckham, Christopher and Weiss, Martin and Golemo, Florian and Honari, Sina and Nowrouzezahrai, Derek and Pal, Christopher},
     journal = {arXiv preprint arXiv:2212.01639},
     year = {2022}
   }

2021

1. iclr

   gradsim: Differentiable simulation for system identification and visuomotor control

   Krishna Murthy Jatavallabhula, Miles Macklin, Florian Golemo, Vikram Voleti, Linda Petrini, Martin Weiss, Breandan Considine, Jérôme Parent-Lévesque, Kevin Xie, Kenny Erleben, and others

   ICLR, 2021

   arXiv Bib

   @article{jatavallabhula2021gradsim,
     title = {gradsim: Differentiable simulation for system identification and visuomotor control},
     author = {Jatavallabhula, Krishna Murthy and Macklin, Miles and Golemo, Florian and Voleti, Vikram and Petrini, Linda and Weiss, Martin and Considine, Breandan and Parent-L{\'e}vesque, J{\'e}r{\o}me and Xie, Kevin and Erleben, Kenny and others},
     journal = {ICLR},
     year = {2021}
   }

2. Sim2real in robotics and automation: Applications and challenges

   Sebastian Höfer, Kostas Bekris, Ankur Handa, Juan Camilo Gamboa, Melissa Mozifian, Florian Golemo, Chris Atkeson, Dieter Fox, Ken Goldberg, John Leonard, and others

   IEEE transactions on automation science and engineering, 2021

   Bib

   @article{hofer2021sim2real,
     title = {Sim2real in robotics and automation: Applications and challenges},
     author = {H{\"o}fer, Sebastian and Bekris, Kostas and Handa, Ankur and Gamboa, Juan Camilo and Mozifian, Melissa and Golemo, Florian and Atkeson, Chris and Fox, Dieter and Goldberg, Ken and Leonard, John and others},
     journal = {IEEE transactions on automation science and engineering},
     volume = {18},
     number = {2},
     pages = {398--400},
     year = {2021},
     publisher = {IEEE}
   }

3. GrowSpace: Learning How to Shape Plants

   Yasmeen Hitti, Ionelia Buzatu, Manuel Del Verme, Mark Lefsrud, Florian Golemo, and Audrey Durand

   arXiv preprint arXiv:2110.08307, 2021

   arXiv Bib

   @article{hitti2021growspace,
     title = {GrowSpace: Learning How to Shape Plants},
     author = {Hitti, Yasmeen and Buzatu, Ionelia and Del Verme, Manuel and Lefsrud, Mark and Golemo, Florian and Durand, Audrey},
     journal = {arXiv preprint arXiv:2110.08307},
     year = {2021}
   }

4. Sequoia: A software framework to unify continual learning research

   Fabrice Normandin, Florian Golemo, Oleksiy Ostapenko, Pau Rodriguez, Matthew D Riemer, Julio Hurtado, Khimya Khetarpal, Ryan Lindeborg, Lucas Cecchi, Timothée Lesort, and others

   arXiv preprint arXiv:2108.01005, 2021

   arXiv Bib

   @article{normandin2021sequoia,
     title = {Sequoia: A software framework to unify continual learning research},
     author = {Normandin, Fabrice and Golemo, Florian and Ostapenko, Oleksiy and Rodriguez, Pau and Riemer, Matthew D and Hurtado, Julio and Khetarpal, Khimya and Lindeborg, Ryan and Cecchi, Lucas and Lesort, Timoth{\'e}e and others},
     journal = {arXiv preprint arXiv:2108.01005},
     year = {2021}
   }

5. iclr

   Latent variable sequential set transformers for joint multi-agent motion prediction

   Roger Girgis, Florian Golemo, Felipe Codevilla, Martin Weiss, Jim Aldon D’Souza, Samira Ebrahimi Kahou, Felix Heide, and Christopher Pal

   arXiv preprint arXiv:2104.00563, 2021

   arXiv Bib

   @article{girgis2021latent,
     title = {Latent variable sequential set transformers for joint multi-agent motion prediction},
     author = {Girgis, Roger and Golemo, Florian and Codevilla, Felipe and Weiss, Martin and D'Souza, Jim Aldon and Kahou, Samira Ebrahimi and Heide, Felix and Pal, Christopher},
     journal = {arXiv preprint arXiv:2104.00563},
     year = {2021}
   }

2020

1. corl

   

   Active Domain Randomization

   Bhairav Mehta, Manfred Diaz, Florian Golemo, Christopher J. Pal, and Liam Paull

   In Proceedings of the Conference on Robot Learning, 30 oct–01 nov 2020

   Abs arXiv Bib PDF

   

   Domain randomization is a popular technique for improving domain transfer, often used in a zero-shot setting when the target domain is unknown or cannot easily be used for training. In this work, we empirically examine the effects of domain randomization on agent generalization. Our experiments show that domain randomization may lead to suboptimal, high-variance policies, which we attribute to the uniform sampling of environment parameters. We propose Active Domain Randomization, a novel algorithm that learns a parameter sampling strategy. Our method looks for the most informative environment variations within the given randomization ranges by leveraging the discrepancies of policy rollouts in randomized and reference environment instances. We find that training more frequently on these instances leads to better overall agent generalization. Our experiments across various physics-based simulated and real-robot tasks show that this enhancement leads to more robust, consistent policies.

   @inproceedings{pmlr-v100-mehta20a,
     title = {Active Domain Randomization},
     author = {Mehta, Bhairav and Diaz, Manfred and Golemo, Florian and Pal, Christopher J. and Paull, Liam},
     booktitle = {Proceedings of the Conference on Robot Learning},
     pages = {1162--1176},
     year = {2020},
     editor = {Kaelbling, Leslie Pack and Kragic, Danica and Sugiura, Komei},
     volume = {100},
     series = {Proceedings of Machine Learning Research},
     month = {30 Oct--01 Nov},
     publisher = {PMLR},
     url = {https://proceedings.mlr.press/v100/mehta20a.html}
   }

2. corl

   Navigation agents for the visually impaired: A sidewalk simulator and experiments

   Martin Weiss, Simon Chamorro, Roger Girgis, Margaux Luck, Samira E Kahou, Joseph P Cohen, Derek Nowrouzezahrai, Doina Precup, Florian Golemo, and Chris Pal

   In Conference on Robot Learning, 2020

   Bib

   @inproceedings{weiss2020navigation,
     title = {Navigation agents for the visually impaired: A sidewalk simulator and experiments},
     author = {Weiss, Martin and Chamorro, Simon and Girgis, Roger and Luck, Margaux and Kahou, Samira E and Cohen, Joseph P and Nowrouzezahrai, Derek and Precup, Doina and Golemo, Florian and Pal, Chris},
     booktitle = {Conference on Robot Learning},
     pages = {1314--1327},
     year = {2020},
     organization = {PMLR}
   }

3. Generating automatic curricula via self-supervised active domain randomization

   Sharath Chandra Raparthy, Bhairav Mehta, Florian Golemo, and Liam Paull

   arXiv preprint arXiv:2002.07911, 2020

   arXiv Bib

   @article{raparthy2020generating,
     title = {Generating automatic curricula via self-supervised active domain randomization},
     author = {Raparthy, Sharath Chandra and Mehta, Bhairav and Golemo, Florian and Paull, Liam},
     journal = {arXiv preprint arXiv:2002.07911},
     year = {2020}
   }

4. Pix2shape: Towards unsupervised learning of 3d scenes from images using a view-based representation

   Sai Rajeswar, Fahim Mannan, Florian Golemo, Jérôme Parent-Lévesque, David Vazquez, Derek Nowrouzezahrai, and Aaron Courville

   International Journal of Computer Vision, 2020

   Bib

   @article{rajeswar2020pix2shape,
     title = {Pix2shape: Towards unsupervised learning of 3d scenes from images using a view-based representation},
     author = {Rajeswar, Sai and Mannan, Fahim and Golemo, Florian and Parent-L{\'e}vesque, J{\'e}r{\o}me and Vazquez, David and Nowrouzezahrai, Derek and Courville, Aaron},
     journal = {International Journal of Computer Vision},
     volume = {128},
     number = {10},
     pages = {2478--2493},
     year = {2020},
     publisher = {Springer US New York}
   }

5. Unsupervised learning of dense visual representations

   Pedro O Pinheiro, Amjad Almahairi, Ryan Y Benmalek, Florian Golemo, and Aaron Courville

   arXiv preprint arXiv:2011.05499, 2020

   arXiv Bib

   @article{pinheiro2020unsupervised,
     title = {Unsupervised learning of dense visual representations},
     author = {Pinheiro, Pedro O and Almahairi, Amjad and Benmalek, Ryan Y and Golemo, Florian and Courville, Aaron},
     journal = {arXiv preprint arXiv:2011.05499},
     year = {2020}
   }

6. journal

   Perspectives on sim2real transfer for robotics: A summary of the r: Ss 2020 workshop

   Sebastian Höfer, Kostas Bekris, Ankur Handa, Juan Camilo Gamboa, Florian Golemo, Melissa Mozifian, Chris Atkeson, Dieter Fox, Ken Goldberg, John Leonard, and others

   arXiv preprint arXiv:2012.03806, 2020

   arXiv Bib

   @article{hofer2020perspectives,
     title = {Perspectives on sim2real transfer for robotics: A summary of the r: Ss 2020 workshop},
     author = {H{\"o}fer, Sebastian and Bekris, Kostas and Handa, Ankur and Gamboa, Juan Camilo and Golemo, Florian and Mozifian, Melissa and Atkeson, Chris and Fox, Dieter and Goldberg, Ken and Leonard, John and others},
     journal = {arXiv preprint arXiv:2012.03806},
     year = {2020}
   }

7. Cunas, a curiosity-driven neural-augmented simulator

   Sharath Chandra Raparthy, Melissa Mozifian, Liam Paull, and Florian Golemo

   In 2nd Workshop on Closing the Reality Gap in Sim2Real Transfer for Robotics. RSS, 2020

   Bib

   @inproceedings{raparthy2020cunascuriosity,
     title = {Cunas, a curiosity-driven neural-augmented simulator},
     author = {Raparthy, Sharath Chandra and Mozifian, Melissa and Paull, Liam and Golemo, Florian},
     booktitle = {2nd Workshop on Closing the Reality Gap in Sim2Real Transfer for Robotics. RSS},
     year = {2020}
   }

2019

1. The ai driving olympics at neurips 2018

   Julian Zilly, Jacopo Tani, Breandan Considine, Bhairav Mehta, Andrea F Daniele, Manfred Diaz, Gianmarco Bernasconi, Claudio Ruch, Jan Hakenberg, Florian Golemo, and others

   In The NeurIPS’18 Competition: From Machine Learning to Intelligent Conversations, 2019

   Bib

   @incollection{zilly2019ai,
     title = {The ai driving olympics at neurips 2018},
     author = {Zilly, Julian and Tani, Jacopo and Considine, Breandan and Mehta, Bhairav and Daniele, Andrea F and Diaz, Manfred and Bernasconi, Gianmarco and Ruch, Claudio and Hakenberg, Jan and Golemo, Florian and others},
     booktitle = {The NeurIPS'18 Competition: From Machine Learning to Intelligent Conversations},
     pages = {37--68},
     year = {2019},
     publisher = {Springer International Publishing Cham}
   }

2. Robo-PlaNet: learning to poke in a day

   Maxime Chevalier-Boisvert, Guillaume Alain, Florian Golemo, and Derek Nowrouzezahrai

   arXiv preprint arXiv:1911.03594, 2019

   arXiv Bib

   @article{chevalier2019robo,
     title = {Robo-PlaNet: learning to poke in a day},
     author = {Chevalier-Boisvert, Maxime and Alain, Guillaume and Golemo, Florian and Nowrouzezahrai, Derek},
     journal = {arXiv preprint arXiv:1911.03594},
     year = {2019}
   }

3. Sidewalk Environment for Visual Navigation

   Martin Weiss, Simon Chamorro, Roger Girgis, Margaux Luck, Samira Kahou, Joseph Cohen, Derek Nowrouzezahrai, Doina Precup, Florian Golemo, and Christopher J Pal

   Zenodo, 2019

   Bib

   @article{weiss2019sidewalk,
     title = {Sidewalk Environment for Visual Navigation},
     author = {Weiss, Martin and Chamorro, Simon and Girgis, Roger and Luck, Margaux and Kahou, Samira and Cohen, Joseph and Nowrouzezahrai, Derek and Precup, Doina and Golemo, Florian and Pal, Christopher J},
     journal = {Zenodo},
     year = {2019}
   }

2018

1. corl

   

   Sim-to-Real Transfer with Neural-Augmented Robot Simulation

   Florian Golemo, Adrien Ali Taiga, Aaron Courville, and Pierre-Yves Oudeyer

   In Proceedings of The 2nd Conference on Robot Learning, 29–31 oct 2018

   Abs Bib PDF

   Despite the recent successes of deep reinforcement learning, teaching complex motor skills to a physical robot remains a hard problem. While learning directly on a real system is usually impractical, doing so in simulation has proven to be fast and safe. Nevertheless, because of the "reality gap," policies trained in simulation often perform poorly when deployed on a real system. In this work, we introduce a method for training a recurrent neural network on the differences between simulated and real robot trajectories and then using this model to augment the simulator. This Neural-Augmented Simulation (NAS) can be used to learn control policies that transfer significantly better to real environments than policies learned on existing simulators. We demonstrate the potential of our approach through a set of experiments on the Mujoco simulator with added backlash and the Poppy Ergo Jr robot. NAS allows us to learn policies that are competitive with ones that would have been learned directly on the real robot.

   @inproceedings{pmlr-v87-golemo18a,
     title = {Sim-to-Real Transfer with Neural-Augmented Robot Simulation},
     author = {Golemo, Florian and Taiga, Adrien Ali and Courville, Aaron and Oudeyer, Pierre-Yves},
     booktitle = {Proceedings of The 2nd Conference on Robot Learning},
     pages = {817--828},
     year = {2018},
     editor = {Billard, Aude and Dragan, Anca and Peters, Jan and Morimoto, Jun},
     volume = {87},
     series = {Proceedings of Machine Learning Research},
     month = {29--31 Oct},
     publisher = {PMLR},
     url = {https://proceedings.mlr.press/v87/golemo18a.html}
   }

2. Pix2Scene: Learning implicit 3D representations from images

   Sai Rajeswar, Fahim Mannan, Florian Golemo, David Vazquez, Derek Nowrouzezahrai, and Aaron Courville

   2018

   Bib

   @article{rajeswar2018pix2scene,
     title = {Pix2Scene: Learning implicit 3D representations from images},
     author = {Rajeswar, Sai and Mannan, Fahim and Golemo, Florian and Vazquez, David and Nowrouzezahrai, Derek and Courville, Aaron},
     year = {2018}
   }

3. How to Train Your Robot. New Environments for Robotic Training and New Methods for Transferring Policies from the Simulator to the Real Robot

   Florian Golemo

   Université de Bordeaux, 2018

   Bib

   @phdthesis{golemo2018train,
     title = {How to Train Your Robot. New Environments for Robotic Training and New Methods for Transferring Policies from the Simulator to the Real Robot},
     author = {Golemo, Florian},
     year = {2018},
     school = {Universit{\'e} de Bordeaux}
   }

4. Duckietown environments for openai gym

   Maxime Chevalier-Boisvert, Florian Golemo, Yanjun Cao, Bhairav Mehta, and Liam Paull

   GitHub repository, 2018

   Bib

   @article{chevalier2018duckietown,
     title = {Duckietown environments for openai gym},
     author = {Chevalier-Boisvert, Maxime and Golemo, Florian and Cao, Yanjun and Mehta, Bhairav and Paull, Liam},
     journal = {GitHub repository},
     year = {2018}
   }

2017

1. A multimodal dataset for object model learning from natural human-robot interaction

   Pablo Azagra, Florian Golemo, Yoan Mollard, Manuel Lopes, Javier Civera, and Ana C Murillo

   In 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2017

   Bib

   @inproceedings{azagra2017multimodal,
     title = {A multimodal dataset for object model learning from natural human-robot interaction},
     author = {Azagra, Pablo and Golemo, Florian and Mollard, Yoan and Lopes, Manuel and Civera, Javier and Murillo, Ana C},
     booktitle = {2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
     pages = {6134--6141},
     year = {2017},
     organization = {IEEE}
   }

2. Home: A household multimodal environment

   Simon Brodeur, Ethan Perez, Ankesh Anand, Florian Golemo, Luca Celotti, Florian Strub, Jean Rouat, Hugo Larochelle, and Aaron Courville

   arXiv preprint arXiv:1711.11017, 2017

   arXiv Bib

   @article{brodeur2017home,
     title = {Home: A household multimodal environment},
     author = {Brodeur, Simon and Perez, Ethan and Anand, Ankesh and Golemo, Florian and Celotti, Luca and Strub, Florian and Rouat, Jean and Larochelle, Hugo and Courville, Aaron},
     journal = {arXiv preprint arXiv:1711.11017},
     year = {2017}
   }

2016

1. A multimodal human-robot interaction dataset

   Pablo Azagra, Yoan Mollard, Florian Golemo, Ana Cristina Murillo, Manuel Lopes, and Javier Civera

   In NIPS 2016, workshop future of interactive learning machines, 2016

   Bib

   @inproceedings{azagra2016multimodal,
     title = {A multimodal human-robot interaction dataset},
     author = {Azagra, Pablo and Mollard, Yoan and Golemo, Florian and Murillo, Ana Cristina and Lopes, Manuel and Civera, Javier},
     booktitle = {NIPS 2016, workshop future of interactive learning machines},
     year = {2016}
   }

2. A multimodal dataset for interactive and incremental learning of object models

   Pablo Azagra, Yoan Mollard, Florian Golemo, Ana C Murillo, Manuel Lopes, and Javier C Civera

   2016

   Bib

   @article{azagra2016multimodam,
     title = {A multimodal dataset for interactive and incremental learning of object models},
     author = {Azagra, Pablo and Mollard, Yoan and Golemo, Florian and Murillo, Ana C and Lopes, Manuel and Civera, Javier C},
     year = {2016}
   }

2015

1. Simulating Morphological Evolution in Large Robot Populations

   Florian Golemo, Marjeta Markovic, Jasper Schoormans, Peter De Jonge, Maik Couwenberg, and Evert Haasdijk

   In Proceedings of the Companion Publication of the 2015 Annual Conference on Genetic and Evolutionary Computation, 2015

   Bib

   @inproceedings{golemo2015simulating,
     title = {Simulating Morphological Evolution in Large Robot Populations},
     author = {Golemo, Florian and Markovic, Marjeta and Schoormans, Jasper and De Jonge, Peter and Couwenberg, Maik and Haasdijk, Evert},
     booktitle = {Proceedings of the Companion Publication of the 2015 Annual Conference on Genetic and Evolutionary Computation},
     pages = {1249--1250},
     year = {2015}
   }

2013

1. The Use of a Service Modeler Together with a PLM Software for the Management of Product-Related Services: A First Use-case-based Approach to Configure Service Components for Product-Related Services

   Christian Zinke, Kyrill Meyer, Florian Golemo, and Lars-Peter Meyer

   In IFIP International Conference on Product Lifecycle Management, 2013

   Bib

   @inproceedings{zinke2013use,
     title = {The Use of a Service Modeler Together with a PLM Software for the Management of Product-Related Services: A First Use-case-based Approach to Configure Service Components for Product-Related Services},
     author = {Zinke, Christian and Meyer, Kyrill and Golemo, Florian and Meyer, Lars-Peter},
     booktitle = {IFIP International Conference on Product Lifecycle Management},
     pages = {449--457},
     year = {2013},
     organization = {Springer Berlin Heidelberg Berlin, Heidelberg}
   }

2. Refinement of the Learning Centre of BMA1

   Florian Golemo

   In MEi: CogSci Conference 2013, Budapest, 2013

   Bib

   @inproceedings{golemo2013refinement,
     title = {Refinement of the Learning Centre of BMA1},
     author = {Golemo, Florian},
     booktitle = {MEi: CogSci Conference 2013, Budapest},
     year = {2013}
   }

2012

1. Künstliche Intelligenz auf Basis von „Baby Machines “: Geschichte und Architektur BMA1

   Florian Golemo

   Dritte Studentenkonferenz Informatik Leipzig 2012 Leipzig, Deutschland, 25. September 2012 Tagungsband, 2012

   Bib

   @article{golemo2012kunstliche,
     title = {K{\"u}nstliche Intelligenz auf Basis von „Baby Machines “: Geschichte und Architektur BMA1},
     author = {Golemo, Florian},
     journal = {Dritte Studentenkonferenz Informatik Leipzig 2012 Leipzig, Deutschland, 25. September 2012 Tagungsband},
     pages = {113},
     year = {2012}
   }

2010

1. Unterstützung von Innovationsprozessen mit IT-Werkzeugen-Der InnoRadar als Teil eines Innovationswerkzeugs

   Florian Golemo

   subs. emis. de, 2010

   Bib

   @article{golemo2010unterstutzung,
     title = {Unterst{\"u}tzung von Innovationsprozessen mit IT-Werkzeugen-Der InnoRadar als Teil eines Innovationswerkzeugs},
     author = {Golemo, Florian},
     journal = {subs. emis. de},
     year = {2010}
   }