Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

On the robustness of action recognition methods in compressed and pixel domain

: Srinivasan, V.; Gul, S.; Bosse, S.; Meyer, J.T.; Schierl, T.; Hellge, C.; Samek, W.


Beghdadi, A. ; Institute of Electrical and Electronics Engineers -IEEE-; European Association for Signal Processing -EURASIP-:
6th European Workshop on Visual Information Processing, EUVIP 2016. Proceedings : October 25-27, 2016, Marseille, France
Piscataway, NJ: IEEE, 2016
ISBN: 978-1-5090-2781-1
ISBN: 978-1-5090-2780-4
ISBN: 978-1-5090-2782-8
European Workshop on Visual Information Processing (EUVIP) <6, 2016, Marseille>
Fraunhofer HHI ()

This paper investigates the robustness of two state-of-theart action recognition algorithms: a pixel domain approach based on 3D convolutional neural networks (C3D) and a compressed domain approach requiring only partial decoding of the video, based on feature description using motion vectors and Fisher vector encoding (MV-FV). We study the robustness of the two algorithms against: (i) quality variations, (ii) changes in video encoding scheme, (iii) changes in resolutions. Experiments are performed on the HMDB51 dataset. Our main findings are that C3D is robust to variations of these parameters while the MV-FV is very sensitive. Hence, we consider C3D as a baseline method for our analysis. We also analyze the reasons behind these different behaviors and discuss their practical implications.