Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Analyzing classifiers: Fisher vectors and deep neural networks

: Bach, S.; Binder, A.; Montavon, G.; Müller, K.-R.; Samek, W.

IEEE Computer Society:
2016 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2016. Proceedings : June 26th - July 1st, 2016, Las Vegas
Piscataway, NJ: IEEE, 2016
ISBN: 978-1-4673-8851-1
Conference on Computer Vision and Pattern Recognition (CVPR) <29, 2016, Las Vegas/Nev.>
Fraunhofer HHI ()

Fisher vector (FV) classifiers and Deep Neural Networks (DNNs) are popular and successful algorithms for solving image classification problems. However, both are generally considered 'black box' predictors as the non-linear transformations involved have so far prevented transparent and interpretable reasoning. Recently, a principled technique, Layer-wise Relevance Propagation (LRP), has been developed in order to better comprehend the inherent structured reasoning of complex nonlinear classification models such as Bag of Feature models or DNNs. In this paper we (1) extend the LRP framework also for Fisher vector classifiers and then use it as analysis tool to (2) quantify the importance of context for classification, (3) qualitatively compare DNNs against FV classifiers in terms of important image regions and (4) detect potential flaws and biases in data. All experiments are performed on the PASCAL VOC 2007 and ILSVRC 2012 data sets.