Control-Aware Prediction Objectives for Autonomous Driving

—Autonomous vehicle software is typically struc- tured as a modular pipeline of individual components (e.g., perception, prediction, and planning) to help separate concerns into interpretable sub-tasks. Even when end-to-end training is possible, each module has its own set of objectives used for safety assurance, sample efficiency, regularization, or interpretability. However, intermediate objectives do not always align with overall system performance. For example, optimizing the likelihood of a trajectory prediction module might focus more on easy-to-predict agents than safety-critical or rare behaviors (e.g., jaywalking). In this paper, we present control- aware prediction objectives (CAPOs), to evaluate the downstream effect of predictions on control without requiring the planner be differentiable. We propose two types of importance weights that weight the predictive likelihood: one using an attention model between agents, and another based on control variation when exchanging predicted trajectories for ground truth trajectories. Experimentally, we show our objectives improve overall system performance in suburban driving scenarios using the CARLA simulator.

[1]  Stephen J. Roberts,et al.  Loss-Calibrated Approximate Inference in Bayesian Neural Networks , 2018, ArXiv.

[2]  M. Ghavamzadeh,et al.  Policy-Aware Model Learning for Policy Gradient Methods , 2020, ArXiv.

[3]  Mengdi Wang,et al.  Model-Based Reinforcement Learning with Value-Targeted Regression , 2020, L4DC.

[4]  Jean Pierre Mercat,et al.  Multi-Head Attention for Multi-Modal Joint Vehicle Motion Forecasting , 2019, 2020 IEEE International Conference on Robotics and Automation (ICRA).

[5]  Daniel Nikovski,et al.  Value-Aware Loss Function for Model-based Reinforcement Learning , 2017, AISTATS.

[6]  Germán Ros,et al.  CARLA: An Open Urban Driving Simulator , 2017, CoRL.

[7]  Lukasz Kaiser,et al.  Attention is All you Need , 2017, NIPS.

[8]  Sanja Fidler,et al.  Learning to Evaluate Perception Models Using Planner-Centric Metrics , 2020, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[9]  Sergey Levine,et al.  Goal-driven dynamics learning via Bayesian optimization , 2017, 2017 IEEE 56th Annual Conference on Decision and Control (CDC).

[10]  S. Savarese,et al.  Goal-Aware Prediction: Learning to Model What Matters , 2020, ICML.

[11]  Galen Clark Haynes,et al.  Beelines: Evaluating Motion Prediction Impact on Self-Driving Safety and Comfort , 2020, ArXiv.

[12]  Marco Pavone,et al.  Rethinking Trajectory Forecasting Evaluation , 2021, ArXiv.

[13]  Wenlong Fu,et al.  Model-based reinforcement learning: A survey , 2018 .

[14]  Roberto Cipolla,et al.  Concrete Problems for Autonomous Vehicle Safety: Advantages of Bayesian Deep Learning , 2017, IJCAI.

[15]  Sammy Omari,et al.  One Thousand and One Hours: Self-driving Motion Prediction Dataset , 2020, CoRL.

[16]  Carl E. Rasmussen,et al.  PILCO: A Model-Based and Data-Efficient Approach to Policy Search , 2011, ICML.

[17]  Sergio Casas,et al.  Perceive, Predict, and Plan: Safe Motion Planning Through Interpretable Semantic Representations , 2020, ECCV.

[18]  Sen Wu,et al.  Understanding and Improving Information Transfer in Multi-Task Learning , 2020, ICLR.

[19]  Priya L. Donti,et al.  Task-based End-to-end Model Learning in Stochastic Optimization , 2017, NIPS.

[20]  Simon Lucey,et al.  Argoverse: 3D Tracking and Forecasting With Rich Maps , 2019, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[21]  Alborz Geramifard,et al.  Reinforcement learning with misspecified model classes , 2013, 2013 IEEE International Conference on Robotics and Automation.

[22]  David Barber,et al.  Bayesian reasoning and machine learning , 2012 .

[23]  Helbing,et al.  Congested traffic states in empirical observations and microscopic simulations , 2000, Physical review. E, Statistical physics, plasmas, fluids, and related interdisciplinary topics.

[24]  Edouard Leurent,et al.  Social Attention for Autonomous Decision-Making in Dense Traffic , 2019, ArXiv.

[25]  Paul Vernaza,et al.  r2p2: A ReparameteRized Pushforward Policy for Diverse, Precise Generative Path Forecasting , 2018, ECCV.

[26]  Jiquan Ngiam,et al.  Large Scale Interactive Motion Forecasting for Autonomous Driving : The Waymo Open Motion Dataset , 2021, 2021 IEEE/CVF International Conference on Computer Vision (ICCV).

[27]  Anca D. Dragan,et al.  Pragmatic Image Compression for Human-in-the-Loop Decision-Making , 2021, NeurIPS.

[28]  Sergey Levine,et al.  Learning Image-Conditioned Dynamics Models for Control of Underactuated Legged Millirobots , 2017, 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[29]  Javier Alonso-Mora,et al.  Planning and Decision-Making for Autonomous Vehicles , 2018, Annu. Rev. Control. Robotics Auton. Syst..

[30]  Justin Dauwels,et al.  Modeling Sensing and Perception Errors towards Robust Decision Making in Autonomous Vehicles , 2020, ArXiv.