Please enable Javascript


Recent ATG R&D publications

Multi-Agent Routing Value Iteration Network

Learning Lane Graph Representations for Motion Forecasting

DAGMapper: Learning to Map by Discovering Lane Topology

Exploiting Sparse Semantic HD Maps for Self-Driving Vehicle Localization

Convolutional Recurrent Network for Road Boundary Extraction

Learning to Localize Through Compressed Binary Maps

Deep Multi-Sensor Lane Detection

Efficient Convolutions for Real-Time Semantic Segmentation of 3D Point Clouds

End-to-End Deep Structured Models for Drawing Crosswalks

Matching Adversarial Networks

Hierarchical Recurrent Attention Networks for Structured Online Maps

DeepRoadMapper: Extracting Road Topology From Aerial Images

R&D Homepage