mlsanityData to 14 Oct 2026

Domain Adaptive Imitation Learning

arXiv 1910.00105 ↗First posted 30 Sep 2019cs.LG cs.AI stat.MLcited by 6

On the map: Offline RL and imitation learning from data → Imitation learning methods

Abstract (arXiv)

We study the question of how to imitate tasks across domains with discrepancies such as embodiment, viewpoint, and dynamics mismatch. Many prior works require paired, aligned demonstrations and an additional RL step that requires environment interactions. However, paired, aligned demonstrations are seldom obtainable and RL procedures are expensive. We formalize the Domain Adaptive Imitation Learning (DAIL) problem, which is a unified framework for imitation learning in the presence of viewpoint, embodiment, and dynamics mismatch. Informally, DAIL is the process of learning how to perform a task optimally, given demonstrations of the task in a distinct domain. We propose a two step approach to DAIL: alignment followed by adaptation. In the alignment step we execute a novel unsupervised MDP alignment algorithm, Generative Adversarial MDP Alignment (GAMA), to learn state and action correspondences from unpaired, unaligned demonstrations. In the adaptation step we leverage the correspondences to zero-shot imitate tasks across domains. To describe when DAIL is feasible via alignment and adaptation, we introduce a theory of MDP alignability. We experimentally evaluate GAMA against baselines in embodiment, viewpoint, and dynamics mismatch scenarios where aligned demonstrations don't exist and show the effectiveness of our approach.

Key facts. Cited by 6 papers in the map (data through 14 Oct 2026). Placed in Imitation learning methods, one of 13 directions in Offline RL and imitation learning from data. Builds on 8 of its 29 references that are in the map. First posted on arXiv 30 Sep 2019.

Builds on

8 of 29 references.

References · 8

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2016Generative Adversarial Imitation Learningcited by 1,597
  3. 2015End-to-End Training of Deep Visuomotor Policiescited by 875
  4. 2017Time-Contrastive Networks: Self-Supervised Learning from Videocited by 459
  5. 2017Imitation from Observation: Learning to Imitate Behaviors from Raw Video via Context Translationcited by 167
  6. 2017Third-Person Imitation Learningcited by 103
  7. 2017Learning Invariant Feature Spaces to Transfer Skills with Reinforcement Learningcited by 85
  8. 2018Cross-Domain Transfer in Reinforcement Learning using Target Apprenticecited by 6

Built on by

Cited by year

Citing papers per year: 2020: 1; 2021: 2; 2022: 1; 2023: 0; 2024: 2.

6 papers.

  1. 2024ODRL: A Benchmark for Off-Dynamics Reinforcement Learningcited by 25
  2. 2021Domain-Robust Visual Imitation Learning with Mutual Information Constraintscited by 20
  3. 2022Learning from Imperfect Demonstrations via Adversarial Confidence Transfercited by 5
  4. 2021Seeing Differently, Acting Similarly: Heterogeneously Observable Imitation Learningcited by 4
  5. 2020HILONet: Hierarchical Imitation Learning from Non-Aligned Observationscited by 1
  6. 2024SENSOR: Imitate Third-Person Expert's Behaviors via Active Sensoringcited by 1

Similar, not cited

Close in wording, no citation either way

  1. 2021Cross-domain Imitation from Observations
  2. 2022Learn what matters: cross-domain imitation learning with task-relevant embeddings
  3. 2019State Alignment-based Imitation Learning
  4. 2020ADAIL: Adaptive Adversarial Imitation Learning
  5. 2023Domain Adaptive Imitation Learning with Visual Observation
  6. 2024Off-Dynamics Reinforcement Learning via Domain Adaptation and Reward Augmented Imitation
  7. 2022Practical Imitation Learning in the Real World via Task Consistency Loss
  8. 2022One-Shot Domain-Adaptive Imitation Learning via Progressive Learning