Imitation Learning from Observations by Minimizing Inverse Dynamics Disagreement

Authors: Chao Yang, Xiaojian Ma, Wenbing Huang, Fuchun Sun, Huaping Liu, Junzhou Huang, Chuang Gan

NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Considerable empirical results on challenging benchmarks indicate that our method attains consistent improvements over other Lf O counterparts.
Researcher Affiliation Collaboration 1 Beijing National Research Center for Information Science and Technology (BNRist), State Key Lab on Intelligent Technology and Systems, Department of Computer Science and Technology, Tsinghua University 2 Center for Vision, Cognition, Learning and Autonomy, Department of Computer Science, UCLA 3 Tencent AI Lab, 4 MIT-IBM Watson AI Lab
Pseudocode Yes Algorithm 1 Inverse-Dynamics-Disagreement-Minimization (IDDM)
Open Source Code No The paper does not contain any statement or link indicating that the source code for the described methodology is publicly available.
Open Datasets Yes To verify the effectiveness of our IDDM, we perform experimental comparisons on seven challenging control tasks, ranging from traditional control to locomotion [8].
Dataset Splits No The paper mentions evaluating
Hardware Specification No The paper does not provide specific details about the hardware (e.g., GPU models, CPU types, memory) used for running the experiments.
Software Dependencies No The paper mentions using "PPO [32]" and "MINE [4, 14]" but does not provide specific version numbers for these or any other software dependencies.
Experiment Setup No Due to the space limit, we defer more detailed specifications of all the evaluated tasks into the supplementary material.