Semantic Predictive Control for Explainable and Efficient Policy Learning
Lists
Only you see this
Abstract
Visual anticipation of ego and object motion over a short time horizons is a key feature of human-level performance in complex environments. We propose a driving policy learning framework that predicts feature representations of future visual inputs; our predictive model infers not only future events but also semantics, which provide a visual explanation of policy decisions. Our Semantic Predictive Control (SPC) framework predicts future semantic segmentation and events by aggregating multi-scale feature maps. A guidance model assists action selection and enables efficient sampling-based optimization. Experiments on multiple simulation environments show that networks which implement SPC can outperform existing model-based reinforcement learning algorithms in terms of data efficiency and total rewards while providing clear explanations for the policy's behavior.
BibTeX
@inproceedings{pan2019semantic,
author = {Xinlei Pan and Xiangyu Chen and Qi-Zhi Cai and John F. Canny and Fisher Yu},
title = {{Semantic Predictive Control for Explainable and Efficient Policy Learning}},
booktitle = {IEEE International Conference on Robotics and Automation},
year = {2019},
url = {https://openreview.net/forum?id=Wc94CW61CT},
}Loading related content...