Perception Datasets for Anomaly Detection in Autonomous Driving: A Survey
This work addresses the need for comprehensive datasets to evaluate anomaly detection methods in autonomous driving, but it is incremental as it surveys existing data rather than introducing new methods or datasets.
This survey tackles the problem of limited data for training deep neural networks in autonomous driving perception systems, which are unreliable with unseen instances, by providing a structured overview and comparison of existing perception datasets for anomaly detection, categorized into real anomalies, synthetic anomalies, and synthetic scenes.
Deep neural networks (DNN) which are employed in perception systems for autonomous driving require a huge amount of data to train on, as they must reliably achieve high performance in all kinds of situations. However, these DNN are usually restricted to a closed set of semantic classes available in their training data, and are therefore unreliable when confronted with previously unseen instances. Thus, multiple perception datasets have been created for the evaluation of anomaly detection methods, which can be categorized into three groups: real anomalies in real-world, synthetic anomalies augmented into real-world and completely synthetic scenes. This survey provides a structured and, to the best of our knowledge, complete overview and comparison of perception datasets for anomaly detection in autonomous driving. Each chapter provides information about tasks and ground truth, context information, and licenses. Additionally, we discuss current weaknesses and gaps in existing datasets to underline the importance of developing further data.