Interpretable Image Classification with Differentiable Prototypes Assignment
Preprint
Abstract
We introduce ProtoPool, an interpretable image classification model with a pool of prototypes shared by the classes. The training is more straightforward than in the existing methods because it does not require the pruning stage. It is obtained by introducing a fully differentiable assignment of prototypes to particular classes. Moreover, we introduce a novel focal similarity function to focus the model on the rare foreground features. We show that ProtoPool obtains state-of-the-art accuracy on the CUB-200-2011 and the Stanford Cars datasets, substantially reducing the number of prototypes. We provide a theoretical analysis of the method and a user study to show that our prototypes are more distinctive than those obtained with competitive methods.Keywords
All Related Versions
This publication has 15 references indexed in Scilit:
- A case-based interpretable deep learning model for classification of mass lesions in digital mammographyNature Machine Intelligence, 2021
- Interpretable, not black-box, artificial intelligence should be used for embryo selectionHuman Reproduction Open, 2021
- Neural Prototype Trees for Interpretable Fine-grained Image Recognition2021
- These do not Look Like Those: An Interpretable Deep Learning Model for Image RecognitionIEEE Access, 2021
- This Looks Like That, Because ... Explaining Prototypes for Interpretable Image Recognition2021
- Concept whitening for interpretable image recognitionNature Machine Intelligence, 2020
- Stop explaining black box machine learning models for high stakes decisions and use interpretable models insteadNature Machine Intelligence, 2019
- The iNaturalist Species Classification and Detection Dataset2018
- The application of two-level attention models in deep convolutional neural network for fine-grained image classification2015
- A neural system analysis of memory storage and retrieval.Psychological Bulletin, 1973