Logo Logo
Help
Contact
Switch Language to German

Rombach, Robin; Esser, Patrick and Ommer, Björn ORCID logoORCID: https://orcid.org/0000-0003-0766-120X (2020): Making Sense of CNNs: Interpreting Deep Representations and Their Invariances with INNs. 16th European Conference on Computer Vision (ECCV), Glasgow, 23.-28. August 2020. Vedaldi, Andrea; Bischof, Horst; Brox, Thomas and Frahm, Jan-Michael (eds.) : In: Computer Vision – ECCV 2020 Proceedings, Vol. 17 Cham: Springer International Publishing. pp. 647-664

Full text not available from 'Open Access LMU'.

Abstract

To tackle increasingly complex tasks, it has become an essential ability of neural networks to learn abstract representations. These task-specific representations and, particularly, the invariances they capture turn neural networks into black box models that lack interpretability. To open such a black box, it is, therefore, crucial to uncover the different semantic concepts a model has learned as well as those that it has learned to be invariant to. We present an approach based on INNs that (i) recovers the task-specific, learned invariances by disentangling the remaining factor of variation in the data and that (ii) invertibly transforms these recovered invariances combined with the model representation into an equally expressive one with accessible semantic concepts. As a consequence, neural network representations become understandable by providing the means to (i) expose their semantic meaning, (ii) semantically modify a representation, and (iii) visualize individual learned semantic concepts and invariances. Our invertible approach significantly extends the abilities to understand black box models by enabling post-hoc interpretations of state-of-the-art networks without compromising their performance.

Actions (login required)

View Item View Item