mlsanityData to 7 Oct 2026

Obstacle Tower: A Generalization Challenge in Vision, Control, and Planning

arXiv 1902.01378 ↗First posted 4 Feb 2019cs.AI cs.LGcited by 54

Not on the map. Why?

Abstract (arXiv)

The rapid pace of recent research in AI has been driven in part by the presence of fast and challenging simulation environments. These environments often take the form of games; with tasks ranging from simple board games, to competitive video games. We propose a new benchmark - Obstacle Tower: a high fidelity, 3D, 3rd person, procedurally generated environment. An agent playing Obstacle Tower must learn to solve both low-level control and high-level planning problems in tandem while learning from pixels and a sparse reward signal. Unlike other benchmarks such as the Arcade Learning Environment, evaluation of agent performance in Obstacle Tower is based on an agent's ability to perform well on unseen instances of the environment. In this paper we outline the environment and provide a set of baseline results produced by current state-of-the-art Deep RL methods as well as human players. These algorithms fail to produce agents capable of performing near human level.

Key facts. Cited by 54 papers in the map (data through 7 Oct 2026). First posted on arXiv 4 Feb 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2020: 1; 2021: 12; 2022: 17; 2023: 9; 2024: 7; 2025: 2; 2026: 6.

54 papers.

Sort:

By area:

  1. 2022MineDojo: Building Open-Ended Embodied Agents with Internet-Scale Knowledgecited by 469
  2. 2022Evolving Curricula with Regret-Based Environment Designcited by 166
  3. 2021A Survey of Zero-shot Generalisation in Deep Reinforcement Learningcited by 164
  4. 2023Human-Timescale Adaptation in an Open-Ended Task Spacecited by 136
  5. 2021Contrastive Behavioral Similarity Embeddings for Generalization in Reinforcement Learningcited by 123
  6. 2022SMACv2: An Improved Benchmark for Cooperative Multi-Agent Reinforcement Learningcited by 118
  7. 2021Scalable Evaluation of Multi-Agent Reinforcement Learning with Melting Potcited by 97
  8. 2021MiniHack the Planet: A Sandbox for Open-Ended Reinforcement Learning Researchcited by 96
  9. 2021Decoupling Value and Policy for Generalization in Reinforcement Learningcited by 81
  10. 2022Exploration via Elliptical Episodic Bonusescited by 51
  11. 2022Contextualize Me -- The Case for Context in Reinforcement Learningcited by 41
  12. 2023Open-World Multi-Task Control Through Goal-Aware Representation Learning and Adaptive Horizon Predictioncited by 39
38 more
  1. 2021CORA: Benchmarks, Baselines, and Metrics as a Platform for Continual Reinforcement Learning Agentscited by 36
  2. 2021Alchemy: A benchmark and analysis toolkit for meta-reinforcement learning agentscited by 35
  3. 2022Look where you look! Saliency-guided Q-networks for generalization in visual Reinforcement Learningcited by 32
  4. 2023On the Importance of Exploration for Generalization in Reinforcement Learningcited by 30
  5. 2023RL-ViGen: A Reinforcement Learning Benchmark for Visual Generalizationcited by 27
  6. 2022MineRL Diamond 2021 Competition: Overview, Results, and Lessons Learnedcited by 26
  7. 2024Odyssey: Empowering Minecraft Agents with Open-World Skillscited by 26
  8. 2023The Generalization Gap in Offline Reinforcement Learningcited by 25
  9. 2021Quantifying the effects of environment and population diversity in multi-agent reinforcement learningcited by 21
  10. 2023A Study of Global and Episodic Bonuses for Exploration in Contextual MDPscited by 19
  11. 2022Avalon: A Benchmark for RL Generalization Using Procedurally Generated Worldscited by 17
  12. 2024DreamCraft: Text-Guided Generation of Functional 3D Environments in Minecraftcited by 17
  13. 2022LECO: Learnable Episodic Count for Task-Specific Intrinsic Rewardcited by 11
  14. 2023Affective Game Computing: A Surveycited by 11
  15. 2024HackAtari: Atari Learning Environments for Robust and Continual Reinforcement Learningcited by 11
  16. 2024A Review of Nine Physics Engines for Reinforcement Learning Researchcited by 9
  17. 2022TropeTwist: Trope-based Narrative Structure Generationcited by 8
  18. 2024Beyond The Rainbow: High Performance Deep Reinforcement Learning on a Desktop PCcited by 7
  19. 2020Demonstration-efficient Inverse Reinforcement Learning in Procedurally Generated Environmentscited by 6
  20. 2022L2Explorer: A Lifelong Reinforcement Learning Assessment Environmentcited by 6
  21. 2022Collaborative Training of Heterogeneous Reinforcement Learning Agents in Environments with Sparse Rewards: What and When to Share?cited by 4
  22. 2022GriddlyJS: A Web IDE for Reinforcement Learningcited by 4
  23. 2022DIAMBRA Arena: a New Reinforcement Learning Platform for Research and Experimentationcited by 4
  24. 2022The Game of Hidden Rules: A New Kind of Benchmark Challenge for Machine Learningcited by 3
  25. 2025Provable Zero-Shot Generalization in Offline Reinforcement Learningcited by 3
  26. 2021Towards robust and domain agnostic reinforcement learning competitionscited by 2
  27. 2022On the Verge of Solving Rocket League using Deep Reinforcement Learning and Sim-to-sim Transfercited by 2
  28. 2023Comparing Reinforcement Learning and Human Learning using the Game of Hidden Rulescited by 2
  29. 2024Autoverse: An Evolvable Game Language for Learning Robust Embodied Agentscited by 2
  30. 2026Children's Intelligence Tests Pose Challenges for MLLMs? KidGym: A 2D Grid-Based Reasoning Benchmark for MLLMscited by 2
  31. 2021Sparse Attention Guided Dynamic Value Estimation for Single-Task Multi-Scene Reinforcement Learningcited by 1
  32. 2021Roominoes: Generating Novel 3D Floor Plans From Existing 3D Roomscited by 1
  33. 2023How much can change in a year? Revisiting Evaluation in Multi-Agent Reinforcement Learningcited by 1
  34. 2024AMaze: An intuitive benchmark generator for fast prototyping of generalizable agentscited by 1
  35. 2025State Combinatorial Generalization In Decision Making With Conditional Diffusion Modelscited by 1
  36. 2021Fast Concept Mapping: The Emergence of Human Abilities in Artificial Neural Networks when Learning Embodied and Self-Supervisedcited by 0
  37. 2022The Sandbox Environment for Generalizable Agent Research (SEGAR)cited by 0
  38. 2026KAGE-Bench: Fast Known-Axis Visual Generalization Evaluation for Reinforcement Learningcited by 0

50 most cited of 54 All 54 (JSON)