SOTAVerified

Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps

2013-12-20Code Available0· sign in to hype

Karen Simonyan, Andrea Vedaldi, Andrew Zisserman

arXivPDF

Code Available — Be the first to reproduce this paper.

Reproduce

Code

Abstract

This paper addresses the visualisation of image classification models, learnt using deep Convolutional Networks (ConvNets). We consider two visualisation techniques, based on computing the gradient of the class score with respect to the input image. The first one generates an image, which maximises the class score [Erhan et al., 2009], thus visualising the notion of the class, captured by a ConvNet. The second technique computes a class saliency map, specific to a given image and class. We show that such maps can be employed for weakly supervised object segmentation using classification ConvNets. Finally, we establish the connection between the gradient-based ConvNet visualisation methods and deconvolutional networks [Zeiler et al., 2013].

Tasks

General ClassificationImage Attributionimage-classificationImage ClassificationInterpretability Techniques for Deep LearningSemantic SegmentationWeakly-Supervised Object Segmentation

Benchmark Results

DatasetModelMetricClaimedVerifiedStatus
CelebASaliencyInsertion AUC score (ArcFace ResNet-101)0.46—Unverified
CUB-200-2011SaliencyInsertion AUC score (ResNet-101)0.66—Unverified
VGGFace2SaliencyInsertion AUC score (ArcFace ResNet-101)0.56—Unverified

Reproductions