Aus Aifbportal
Wechseln zu:Navigation, Suche

3D Voxel Benchmark for Anomaly Detection in Autonomous Driving

Informationen zur Arbeit

Abschlussarbeitstyp: Bachelor, Master
Betreuer: Daniel Bogdoll
Forschungsgruppe: Angewandte Technisch-Kognitive Systeme

Archivierungsnummer: 4945
Abschlussarbeitsstatus: Offen
Beginn: 06. September 2022
Abgabe: unbekannt

Weitere Informationen

We don’t have autonomous vehicles yet because they’re bad at dealing with the crazy things we see on the roads. Thus, anomaly detection is particularly important to detect the unknown. However, there exists no strong benchmark to compare deep learning anomaly detection methods. In this thesis, you will develop a multimodal benchmark with hundreds of anomalies and ground truth in the voxel space.

Ausschreibung: Download (pdf)