Here you can find useful links to the presentations, code and papers of the finalists in previous MineRL competitions, as well as other publications and projects that use MineRL.
To see all papers that cite MineRL, check Google Scholar. You can also create alerts there to get notified whenever a new citation appears.
If you want to add your paper/project here, do not hesitate to create a pull request in the main repository!
- MineRL 2019 - Finalists presentations at NeurIPS 2019
- MineRL 2019 - 1st place winners presentation, longer one (slides in English, talk in Russian)
- MineRL 2020 - Round 1 finalists presentations at NeurIPS 2020
- MineRL 2020 - Round 2 finalists presentations at Microsoft AI and Gaming Research Summit 2021
- MineRL: A Large-Scale Dataset of Minecraft Demonstrations
- The MineRL 2019 Competition on Sample Efficient Reinforcement Learning using Human Priors
- Retrospective Analysis of the 2019 MineRL Competition on Sample Efficient Reinforcement Learning
- The MineRL 2020 Competition on Sample Efficient Reinforcement Learning using Human Priors
- Towards robust and domain agnostic reinforcement learning competitions: MineRL 2020
- 1st place: paper.
- 2nd place: paper, code.
- 3rd place: paper, code.
- 4th place: code.
- 5th place: paper, code.
- PiCoEDL: Discovery and Learning of Minecraft Navigation Goals from Pixels and Coordinates (CVPR Embodied AI Workshop, 2021)
- Universal Value Iteration Networks: When Spatially-Invariant Is Not Universal (AAAI, 2020)
- Multi-task curriculum learning in a complex, visual, hard-exploration domain: Minecraft
- Follow up paper from the #1 team in 2019 (obtains diamond): paper, code.
- Align-RUDDER: Learning From Few Demonstrations by Reward Redistribution (obtains diamond): paper, code.
- IGLU: Interactive Grounded Language Understanding in a Collaborative Environment: paper (NEURIPS, 2022) Other -----