We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CV

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computer Vision and Pattern Recognition

Title: Analyzing Classifiers: Fisher Vectors and Deep Neural Networks

Abstract: Fisher Vector 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 data set.
Comments: 17 pages (10 main document + references , 7 appendix) 1 Table 7 Figures 1 Algorithm submitted to CVPR on 06/11/2025
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Cite as: arXiv:1512.00172 [cs.CV]
  (or arXiv:1512.00172v1 [cs.CV] for this version)

Submission history

From: Sebastian Bach [view email]
[v1] Tue, 1 Dec 2015 08:14:11 GMT (7478kb,D)

Link back to: arXiv, form interface, contact.