IntentNet: Learning to Predict Intention from Raw Sensor Data

Abstract

In order to plan a safe maneuver, self-driving vehicles need to understand the intent of other traffic participants. We define intent as a combination of discrete high-level behaviors as well as continuous trajectories describing future motion. In this paper, we develop a one-stage detector and forecaster that exploits both 3D point clouds produced by a LiDAR sensor as well as dynamic maps of the environment. Our multi-task model achieves better accuracy than the respective separate modules while saving computation, which is critical to reducing reaction time in self-driving applications.

CoRL 2018

Cited by in corpus (32)

• TNT: Target-driveN Trajectory Prediction
• Argoverse: 3D Tracking and Forecasting with Rich Maps
• MultiPath: Multiple Probabilistic Anchor Trajectory Hypotheses for Behavior Prediction
• Semantics for Robotic Mapping, Perception and Interaction: A Survey
• Spatially-Aware Graph Neural Networks for Relational Behavior Forecasting from Sensor Data
• VectorNet: Encoding HD Maps and Agent Dynamics from Vectorized Representation
• Learning Lane Graph Representations for Motion Forecasting
• BEV-Seg: Bird's Eye View Semantic Segmentation Using Geometry and Semantic Point Cloud
• Joint Interaction and Trajectory Prediction for Autonomous Driving using Graph Neural Networks
• Auto4D: Learning to Label 4D Objects from Sequential Point Clouds
• What-If Motion Prediction for Autonomous Driving
• Rules of the Road: Predicting Driving Behavior with a Convolutional Model of Semantic Interactions
• Deep Learning for Vision-based Prediction: A Survey
• V2VNet: Vehicle-to-Vehicle Communication for Joint Perception and Prediction
• Discrete Residual Flow for Probabilistic Pedestrian Behavior Prediction
• Probabilistic Multi-modal Trajectory Prediction with Lane Attention for Autonomous Vehicles
• DSDNet: Deep Structured self-Driving Network
• An LSTM Approach to Temporal 3D Object Detection in LiDAR Point Clouds
• STINet: Spatio-Temporal-Interactive Network for Pedestrian Detection and Trajectory Prediction
• PePScenes: A Novel Dataset and Baseline for Pedestrian Action Prediction in 3D
• PnPNet: End-to-End Perception and Prediction with Tracking in the Loop
• TPCN: Temporal Point Cloud Networks for Motion Forecasting
• Planning Robot Motion using Deep Visual Prediction
• End-to-end Interpretable Neural Motion Planner
• Safety-Oriented Pedestrian Motion and Scene Occupancy Forecasting
• TrafficSim: Learning to Simulate Realistic Multi-Agent Behaviors
• Investigating the Effect of Sensor Modalities in Multi-Sensor Detection-Prediction Models
• Just Label What You Need: Fine-Grained Active Selection for Perception and Prediction through Partially Labeled Scenes
• ParkPredict: Motion and Intent Prediction of Vehicles in Parking Lots
• Open-set Intersection Intention Prediction for Autonomous Driving
• DAGMapper: Learning to Map by Discovering Lane Topology
• Convolutional Recurrent Network for Road Boundary Extraction