AVOID: The Adverse Visual Conditions Dataset with Obstacles for Driving Scene Understanding

TL;DR

The paper introduces AVOID, a comprehensive dataset for real-time obstacle detection in adverse driving conditions, including diverse visual and sensor data, to improve autonomous vehicle scene understanding.

Contribution

It provides a new large-scale dataset with multi-modal data capturing unexpected obstacles under adverse conditions, supporting various perception tasks and benchmarking.

Findings

High performance obstacle detection results on the dataset

Ablation studies on multi-task perception networks

Enhanced understanding of adverse condition impacts

Abstract

Understanding road scenes for visual perception remains crucial for intelligent self-driving cars. In particular, it is desirable to detect unexpected small road hazards reliably in real-time, especially under varying adverse conditions (e.g., weather and daylight). However, existing road driving datasets provide large-scale images acquired in either normal or adverse scenarios only, and often do not contain the road obstacles captured in the same visual domain as for the other classes. To address this, we introduce a new dataset called AVOID, the Adverse Visual Conditions Dataset, for real-time obstacle detection collected in a simulated environment. AVOID consists of a large set of unexpected road obstacles located along each path captured under various weather and time conditions. Each image is coupled with the corresponding semantic and depth maps, raw and semantic LiDAR data, and waypoints, thereby supporting most visual perception tasks. We benchmark the results on high-performing real-time networks for the obstacle detection task, and also propose and conduct ablation studies using a comprehensive multi-task network for semantic segmentation, depth and waypoint prediction tasks.