conference paper


Conference/ProceedingsIEEE International Conference on Advanced Video and Signal-Based Surveillance
Start date23.08.2016
End date26.08.2016
AddressColorado Springs, CO, USA
Pages278-285
Author(s)Erik Bochinski, Volker Eiselein, Thomas Sikora
TitleTraining a Convolutional Neural Network for Multi-Class Object Detection Using Solely Virtual World Data
AbstractConvolutional 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.
Key wordsDeep Learning, Convolutional Neural Network, Object Detection, Dataset Synthesis, MOCATDataset
NoteElectronic ISBN: 978-1-5090-3811-4
Print on Demand(PoD) ISBN: 978-1-5090-3812-1
DOI: 10.1109/AVSS.2016.7738056
File1491Bochinski2016.pdf

[BibTeX]