Neural Prototype Trees for Interpretable Fine-grained Image Recognition

TL;DR

ProtoTree introduces an interpretable deep learning model combining prototype learning and decision trees for fine-grained image recognition, enabling both global and local explanations while maintaining competitive accuracy.

Contribution

It presents the Neural Prototype Tree (ProtoTree), a novel intrinsically interpretable model that integrates prototype learning with decision trees for fine-grained image classification.

Findings

Small trees with 8 prototypes classify 200 bird species.

Ensemble of 5 ProtoTrees achieves competitive accuracy.

Pruning maintains accuracy while reducing model complexity.

Abstract

Prototype-based methods use interpretable representations to address the black-box nature of deep learning models, in contrast to post-hoc explanation methods that only approximate such models. We propose the Neural Prototype Tree (ProtoTree), an intrinsically interpretable deep learning method for fine-grained image recognition. ProtoTree combines prototype learning with decision trees, and thus results in a globally interpretable model by design. Additionally, ProtoTree can locally explain a single prediction by outlining a decision path through the tree. Each node in our binary tree contains a trainable prototypical part. The presence or absence of this learned prototype in an image determines the routing through a node. Decision making is therefore similar to human reasoning: Does the bird have a red throat? And an elongated beak? Then it's a hummingbird! We tune the accuracy-interpretability trade-off using ensemble methods, pruning and binarizing. We apply pruning without sacrificing accuracy, resulting in a small tree with only 8 learned prototypes along a path to classify a bird from 200 species. An ensemble of 5 ProtoTrees achieves competitive accuracy on the CUB-200- 2011 and Stanford Cars data sets. Code is available at https://github.com/M-Nauta/ProtoTree