ROAD: The ROad event Awareness Dataset for Autonomous Driving

Abstract

Humans drive in a holistic fashion which entails, in particular, understanding dynamic road events and their evolution. Injecting these capabilities in autonomous vehicles can thus take situational awareness and decision making closer to human-level performance. To this purpose, we introduce the ROad event Awareness Dataset (ROAD) for Autonomous Driving, to our knowledge the first of its kind. ROAD is designed to test an autonomous vehicle's ability to detect road events, defined as triplets composed by an active agent, the action(s) it performs and the corresponding scene locations. ROAD comprises videos originally from the Oxford RobotCar Dataset annotated with bounding boxes showing the location in the image plane of each road event. We benchmark various detection tasks, proposing as a baseline a new incremental algorithm for online road event awareness termed 3D-RetinaNet. We also report the performance on the ROAD tasks of Slowfast and YOLOv5 detectors, as well as that of the winners of the ICCV2021 ROAD challenge, which highlight the challenges faced by situation awareness in autonomous driving. ROAD is designed to allow scholars to investigate exciting tasks such as complex (road) activity detection, future event anticipation and continual learning. The dataset is available at https://github.com/gurkirt/road-dataset; the baseline can be found at https://github.com/gurkirt/3D-RetinaNet.

arXiv comment29 pages, accepted at TPAMI

References in corpus (13)

• Two-Stream Convolutional Networks for Action Recognition in Videos
• UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild
• The Kinetics Human Action Video Dataset
• The ApolloScape Open Dataset for Autonomous Driving and its Application
• TraPHic: Trajectory Prediction in Dense and Heterogeneous Traffic Using Weighted Interactions
• Charades-Ego: A Large-Scale Dataset of Paired Third and First Person Videos
• Human Action Localization with Sparse Spatial Supervision
• A Better Baseline for AVA
• D${}^2$-City: A Large-Scale Dashcam Video Dataset of Diverse Traffic Scenarios
• Occam's razor is insufficient to infer the preferences of irrational agents
• Consistent Optimization for Single-Shot Object Detection
• Multi-modal panoramic 3D outdoor datasets for place categorization
• Incremental Tube Construction for Human Action Detection

Cited by in corpus (6)

• Motion Planning for Autonomous Driving: The State of the Art and Future Perspectives
• Explainable AI for Safe and Trustworthy Autonomous Driving: A Systematic Review
• ROAD-R: The Autonomous Driving Dataset with Logical Requirements
• AssistTaxi: A Comprehensive Dataset for Taxiway Analysis and Autonomous Operations
• Monitoring and Adapting the Physical State of a Camera for Autonomous Vehicles
• An object detection approach for lane change and overtake detection from motion profiles