Generative Visual Rationales

Interpretability and small labelled datasets are key issues in the practical application of deep learning, particularly in areas such as medicine. In this paper, we present a semi-supervised technique that addresses both these issues by leveraging large unlabelled datasets to encode and decode images into a dense latent representation. Using chest radiography as an example, we apply this encoder to other labelled datasets and apply simple models to the latent vectors to learn algorithms to identify heart failure. For each prediction, we generate visual rationales by optimizing a latent representation to minimize the prediction of disease while constrained by a similarity measure in image space. Decoding the resultant latent representation produces an image without apparent disease. The difference between the original decoding and the altered image forms an interpretable visual rationale for the algorithm's prediction on that image. We also apply our method to the MNIST dataset and compare the generated rationales to other techniques described in the literature.

[1]  Anil A. Bharath,et al.  Inverting the Generator of a Generative Adversarial Network , 2016, IEEE Transactions on Neural Networks and Learning Systems.

[2]  Abhishek Kumar,et al.  Improved Semi-supervised Learning with GANs using Manifold Invariances , 2017, NIPS 2017.

[3]  Sara Kennedy,et al.  Ability of physicians to diagnose congestive heart failure based on chest X-ray. , 2011, The Journal of emergency medicine.

[4]  David Lopez-Paz,et al.  Optimizing the Latent Space of Generative Networks , 2017, ICML.

[5]  Andrew Zisserman,et al.  Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps , 2013, ICLR.

[6]  H. Krum,et al.  B-Type Natriuretic Peptide Testing and the Accuracy of Heart Failure Diagnosis in the Emergency Department , 2010, Circulation. Heart failure.

[7]  Bolei Zhou,et al.  Learning Deep Features for Discriminative Localization , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[8]  Rob Fergus,et al.  Visualizing and Understanding Convolutional Networks , 2013, ECCV.

[9]  Yoshua Bengio,et al.  Generative Adversarial Nets , 2014, NIPS.

[10]  Ankur Taly,et al.  Axiomatic Attribution for Deep Networks , 2017, ICML.

[11]  Trevor Darrell,et al.  Generating Visual Explanations , 2016, ECCV.

[12]  Carlos Guestrin,et al.  "Why Should I Trust You?": Explaining the Predictions of Any Classifier , 2016, ArXiv.

[13]  Sepp Hochreiter,et al.  Self-Normalizing Neural Networks , 2017, NIPS.

[14]  Aaron C. Courville,et al.  Improved Training of Wasserstein GANs , 2017, NIPS.

[15]  Léon Bottou,et al.  Wasserstein Generative Adversarial Networks , 2017, ICML.