End-to-End Differentiable Adversarial Imitation Learning

By Nir Baram, Oron Anschel, Itai Caspi, Shie Mannor. 2017

114 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

A Framework for Data-Driven Robotics (2019, background): “Reward learning: Learning reward functions using inverse RL dates back to Ng and Russell [35] and has achieved tremendous success recently [36]–[42].”
An Algorithmic Perspective on Imitation Learning (2018, methodology): “Recent work by Baram et al. [2017] extended GAIL to the modelbased approach.”
Discriminator-Actor-Critic: Addressing Sample Inefficiency and Reward Bias in Adversarial Imitation Learning (2018, background): “More recent work extends the framework by improving on stability and robustness (Wang et al., 2017; Kim & Park, 2018) and making connections to model-based imitation learning (Baram et al., 2017).”
Task-Relevant Adversarial Imitation Learning (2019, background): “Although many variants are introduced in the literature [Li et al., 2017, Fu et al., 2018, Merel et al., 2017, Zhu et al., 2018, Baram et al., 2017], making GAIL work for high-dimensional input spaces, particularly raw pixels, remains a challenge.”
Scaling data-driven robotics with reward sketching and batch reinforcement learning (2019, methodology): “the robot or the environment, and, thus, no domain gap. Reward learning: Learning reward functions using inverse RL dates back to Ng and Russell [35] and has achieved tremendous success recently [36]–[42]. This class of methods works best when applied to states or well-engineered features. Making it”
Universal Planning Networks (2018, background): “Learning representations capable of providing metric based rewards naturally relates to inverse reinforcement learning (IRL) (Ng & Russell, 2000; Abbeel & Ng, 2004; Finn et al., 2016a; Ho & Ermon, 2016; Baram et al., 2017) and reward shaping (Ng et al.”
Generative Adversarial Self-Imitation Learning (2018, methodology): “Since MGAIL has been shown to be more sample-efficient, we expect that a model-based counterpart of GASIL would also improve the performance.”
Reinforcement Learning on Web Interfaces Using Workflow-Guided Exploration (2018, methodology): “When only expert demonstrations are available, inverse reinforcement learning methods (Abbeel & Ng, 2004; Ziebart et al., 2008; Finn et al., 2016; Ho & Ermon, 2016; Baram et al., 2017) infer a reward function from the demonstrations without using reinforcement signals from the environment.”
Random Expert Distillation: Imitation Learning via Expert Policy Support Estimation (2019, methodology): “Generative Adversarial Imitation Learning (GAIL) Ho & Ermon (2016); Baram et al. (2017) connects IRL to the general framework of Generative Adversarial Networks (GANs) Goodfellow et al. 1Department of Electrical and Electronic Engineering, Imperial College London, SW7 2BT, United Kingdom ar X iv…”
A Survey of Imitation Learning Methods, Environments and Metrics (2024, methodology): “In learning from demonstration hybrid methods, Baram et al. develop a method that optimises the forward dynamics model [7].”
Positive-Unlabeled Reward Learning (2019, background): “Generative Adversarial Imitation Learning (GAIL; Ho & Ermon, 2016), has accrued a wide range of variants including Li et al. (2017), Fu et al. (2017), and Baram et al. (2017).”
r2p2: A ReparameteRized Pushforward Policy for Diverse, Precise Generative Path Forecasting (2018, background): “A similar idea was explored for general imitation learning in [6].”
Hierarchical Approaches for Reinforcement Learning in Parameterized Action Space (2018, methodology): “This trick has been applied to the RL setting as well, including imitation learning (Baram et al. 2017) and multiagent RL (Mordatch and Abbeel 2017).”
Imitation Learning: A Survey of Learning Methods, Environments and Metrics (2024, unknown): “Recent work in imitation learning has focused on taking inspiration from adversarial and dynamics methods [3, 17, 21, 22, 55, 68–71].”