Inverting Visual Representations with Convolutional Networks
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016
Abstract: Feature representations, both hand-designed and learned ones, are often hard to analyze and interpret, even when they are extracted from visual data. We propose a new approach to study image representations by inverting them with an up-convolutional neural network. We apply the method to shallow representations (HOG, SIFT, LBP), as well as to deep networks. For shallow representations our approach provides significantly better reconstructions than existing methods, revealing that there is surprisingly rich information contained in these features when combined with a strong prior. Inverting a deep network trained on ImageNet provides several insights into the properties of the feature representation learned by the network. Most strikingly, the colors and the rough contours of an image can be reconstructed from activations in higher network layers and even from the predicted class probabilities.
DownloadsImages and movies
BibTex reference
@InProceedings{DB16, author = "A. Dosovitskiy and T.Brox", title = "Inverting Visual Representations with Convolutional Networks", booktitle = "IEEE Conference on Computer Vision and Pattern Recognition (CVPR)", month = " ", year = "2016", note = "arXiv:1506.02753", url = "http://lmb.informatik.uni-freiburg.de/Publications/2016/DB16" }