TU Berlin

Fachgebiet NachrichtenübertragungWiss. Veröffentlichungen

Inhalt des Dokuments

zur Navigation

Wissenschaftliche Veröffentlichungen

Training a Convolutional Neural Network for Multi-Class Object Detection Using Solely Virtual World Data
Zitatschlüssel 1491Bochinski2016
Autor Erik Bochinski and Volker Eiselein and Thomas Sikora
Buchtitel IEEE International Conference on Advanced Video and Signal-Based Surveillance
Seiten 278–285
Jahr 2016
Adresse Colorado Springs, CO, USA
Monat aug
Notiz Electronic ISBN: 978-1-5090-3811-4 Print on Demand(PoD) ISBN: 978-1-5090-3812-1 DOI: 10.1109/AVSS.2016.7738056
Zusammenfassung Convolutional neural networks are a popular choice for current object detection and classification systems. Their performance improves constantly but for effective training, large, hand-labeled datasets are required. We address the problem of obtaining customized, yet large enough datasets for CNN training by synthesizing them in a virtual world, thus eliminating the need for tedious human interaction for ground truth creation. We developed a CNN-based multi-class detection system that was trained solely on virtual world data and achieves competitive results compared to state-of-the-art detection systems.
Link zur Publikation Download Bibtex Eintrag

Navigation

Direktzugang

Schnellnavigation zur Seite über Nummerneingabe