The First Workshop on

Out-of-Label Hazards

in Autonomous Driving

Winter Conference on Applications of Computer Vision (WACV) 2025 - Tucson, Arizona

Workshop day: March 4th, 2025 (PST)

Where: Salon F, Room F07, JW Marriott Starpass

In recent years, the field of autonomous vehicles has rapidly expanded, driven by advancements in AI, machine learning, and sensor technologies that promise to enhance safety, reduce traffic accidents, and improve mobility. Avoiding hazards is essential from the safety perspective for a robust system, and there is room for further improvement. A robust autonomous vehicle system should be able to identify and appropriately respond to unexpected or previously unseen hazards to avoid accidents. In this workshop, we encourage researchers to address the limitations of current approaches for avoiding hazards on the road by proposing new algorithms and systems that advance the field, in particular with solutions inspired by novelty-adjacent areas such as Anomaly Detection, Open-Set Recognition, Open Vocabulary, and Domain Adaptation.

Topics

The list of topics includes but is not limited to the following:

Detecting, recognizing, predicting, and avoiding out-of-label hazards in autonomous driving.
Detecting, recognizing, predicting activity understanding in autonomous driving
Handling low-resolution hazards.
New datasets and metrics for autonomous driving.
Vision systems for autonomous driving.
Vision-language models for autonomous driving.
HRI: Human Factors in Autonomous Driving
Explainable AI (XAI) in autonomous driving.