Learning Human Driving Behaviors with Sequential Causal Imitation Learning

Authors: Kangrui Ruan, Xuan Di4583-4592

AAAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our methods are evaluated on a synthetic dataset and a real-world highway driving dataset, both demonstrating that the proposed procedure significantly outperforms noncausal imitation learning methods.We demonstrate the superiority of our proposed method over non-causal baselines by conducting experiments on a car-following dataset and a real-world highway dataset.
Researcher Affiliation Academia Kangrui Ruan, Xuan Di Columbia University, New York, NY, USA kr2910@columbia.edu, sharon.di@columbia.edu
Pseudocode Yes Algorithm 1: Finding π(a|scausal) by GAIL
Open Source Code No The paper does not contain any statement about releasing open-source code or provide a link to a code repository for the described methodology.
Open Datasets Yes The public NGSIM data records the real-world human driving trajectories for US Highway 101, and Interstate 80 Freeway (Alexiadis et al. 2004).
Dataset Splits No The paper mentions using a 'synthetic dataset' and a 'real-world highway driving dataset, Next Generation Simulation (NGSIM)' and that 'expert demonstrations are generated by running PPO-clip'. However, it does not provide specific percentages, sample counts, or explicit instructions for how the datasets were split into training, validation, and test sets.
Hardware Specification No The paper does not provide any specific details about the hardware (e.g., GPU models, CPU types, memory) used to run the experiments.
Software Dependencies No The paper does not list specific software dependencies with version numbers (e.g., library names, frameworks, or programming language versions like 'Python 3.8' or 'PyTorch 1.9').
Experiment Setup No The paper describes the overall experimental approach, including the use of GAIL and PPO, and mentions 'ϵ is a hyper-parameter to control the clip value' in the PPO objective. However, it does not provide concrete numerical values for key hyperparameters such as learning rate, batch size, number of epochs, or other optimizer settings necessary for explicit reproduction of the experimental setup.