Multimodal Trajectory Predictions for Autonomous Driving using Deep Convolutional Networks

Henggang Cui,Vladan Radosavljevic,Fang-Chieh Chou,Tsung-Han Lin,Thi Nguyen,Tzu-Kuo Huang,Jeff Schneider,Nemanja Djuric

DOI: https://doi.org/10.48550/arXiv.1809.10732

2019-03-01

Abstract:Autonomous driving presents one of the largest problems that the robotics and artificial intelligence communities are facing at the moment, both in terms of difficulty and potential societal impact. Self-driving vehicles (SDVs) are expected to prevent road accidents and save millions of lives while improving the livelihood and life quality of many more. However, despite large interest and a number of industry players working in the autonomous domain, there still remains more to be done in order to develop a system capable of operating at a level comparable to best human drivers. One reason for this is high uncertainty of traffic behavior and large number of situations that an SDV may encounter on the roads, making it very difficult to create a fully generalizable system. To ensure safe and efficient operations, an autonomous vehicle is required to account for this uncertainty and to anticipate a multitude of possible behaviors of traffic actors in its surrounding. We address this critical problem and present a method to predict multiple possible trajectories of actors while also estimating their probabilities. The method encodes each actor's surrounding context into a raster image, used as input by deep convolutional networks to automatically derive relevant features for the task. Following extensive offline evaluation and comparison to state-of-the-art baselines, the method was successfully tested on SDVs in closed-course tests.

Robotics,Computer Vision and Pattern Recognition,Machine Learning

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to predict the multi - modal trajectories of surrounding traffic participants (such as other vehicles, pedestrians, etc.) in self - driving vehicles (SDVs) and estimate the probabilities of these trajectories. Specifically, the paper aims to develop a method that can predict multiple possible future trajectories while taking into account the high uncertainty and complexity of traffic behavior, thereby improving the safety and efficiency of self - driving vehicles. This involves capturing the context information of each traffic participant in a dynamic environment and using deep convolutional networks to automatically extract relevant features to achieve the prediction of multiple possible behaviors. Through this method, the paper attempts to overcome the limitations of existing unimodal prediction models, especially the poor performance of these models when dealing with long - term predictions and their inability to fully capture the multi - modal nature of traffic scenes.

Multimodal Trajectory Predictions for Autonomous Driving using Deep Convolutional Networks

Enhanced Multimodal Trajectory Prediction for Autonomous Vehicles Using Advanced Diffusion Model Techniques

Multimodal Trajectory Prediction for Autonomous Driving on Unstructured Roads using Deep Convolutional Network

Deep Kinematic Models for Kinematically Feasible Vehicle Trajectory Predictions

Shared Cross-Modal Trajectory Prediction for Autonomous Driving

Multi-modal Trajectory Prediction for Autonomous Driving with Semantic Map and Dynamic Graph Attention Network

Trajectory Prediction for Human-driven Vehicles Based on a Combined Method of Deep Neural Network and Driving Risk Map

Deep learning-based vehicle trajectory prediction based on generative adversarial network for autonomous driving applications

Multimodal Trajectory Predictions for Autonomous Driving without a Detailed Prior Map

Probabilistic multi-modal expected trajectory prediction based on LSTM for autonomous driving

Multiple Contextual Cues Integrated Trajectory Prediction for Autonomous Driving

Vehicle Trajectory Prediction with Multimodal and Dynamics-Aware Interaction Neural Networks

Vehicle Trajectory Prediction for Automated Driving Based on Temporal Convolution Networks

Graph-Based Interaction-Aware Multimodal 2D Vehicle Trajectory Prediction using Diffusion Graph Convolutional Networks

Multimodal Trajectory Prediction for Diverse Vehicle Types in Autonomous Driving with Heterogeneous Data and Physical Constraints

A Cognitive-Based Trajectory Prediction Approach for Autonomous Driving

An Improved Multimodal Trajectory Prediction Method Based on Deep Inverse Reinforcement Learning

Map-Adaptive Multimodal Trajectory Prediction via Intention-Aware Unimodal Trajectory Predictors

Multiagent Multimodal Trajectory Prediction in Urban Traffic Scenarios Using a Neural Network-Based Solution

Human Observation-Inspired Trajectory Prediction for Autonomous Driving in Mixed-Autonomy Traffic Environments

Interactive Trajectory Prediction Using a Driving Risk Map-Integrated Deep Learning Method for Surrounding Vehicles on Highways