Benutzer: Gast  Login
Titel:

Transfer Learning from Simulated to Real Scenes for Monocular 3D Object Detection

Dokumenttyp:
Konferenzbeitrag
Art des Konferenzbeitrags:
Textbeitrag / Aufsatz
Autor(en):
Mohamed, Sondos; Zimmer, Walter; Greer, Ross; Alaaeldin Ghita, Ahmed; Castrillón-Santana, Modesto; Trivedi, Mohan M.; Knoll, Alois C.; Carta, Salvatore Mario; Marras, Mirko
Seitenangaben Beitrag:
18
Abstract:
Accurately detecting 3D objects from monocular images in dynamic roadside scenarios remains a challenging problem due to varying camera perspectives and unpredictable scene conditions. This paper introduces a two-stage training strategy to address these challenges. Our approach initially trains a model on the large-scale synthetic dataset, RoadSense3D, which offers a diverse range of scenarios for robust feature learning. Subsequently, we fine-tune the model on a combination of real-world datase...     »
Stichworte:
Roadside Perception, Autonomous Driving, Dataset, Monocular 3D Perception, 3D Object Detection, Synthetic Dataset, Transfer Learning
Dewey-Dezimalklassifikation:
000 Informatik, Wissen, Systeme
Herausgeber:
ECVA
Kongress- / Buchtitel:
Proceedings of the 18th European Conference on Computer Vision ECCV 2024
Ausrichter der Konferenz:
Springer
Datum der Konferenz:
30 September 2024
Verlag / Institution:
Springer-Verlag
Publikationsdatum:
30.09.2024
Jahr:
2024
Quartal:
4. Quartal
Jahr / Monat:
2024-09
Monat:
Sep
Seiten:
19
Nachgewiesen in:
Scopus; Web of Science
Reviewed:
ja
Sprache:
en
Erscheinungsform:
WWW
TUM Einrichtung:
TUM School of Computation, Information and Technology
Format:
Text
CC-Lizenz:
by, http://creativecommons.org/licenses/by/4.0
 BibTeX