This Looks Better than That: Better Interpretable Models with ProtoPNeXt
This work addresses the challenge of applying interpretable models to new datasets in computer vision, though it appears incremental as it builds on existing ProtoPNet methods.
The paper tackles the difficulty of training prototypical-part models by introducing ProtoPNeXt, a framework for integrating their components, and shows that applying Bayesian hyperparameter tuning and an angular prototype similarity metric achieves new state-of-the-art accuracy on CUB-200, with accuracy changes between +1.3% and -1.5% when optimizing for both accuracy and interpretability.
Prototypical-part models are a popular interpretable alternative to black-box deep learning models for computer vision. However, they are difficult to train, with high sensitivity to hyperparameter tuning, inhibiting their application to new datasets and our understanding of which methods truly improve their performance. To facilitate the careful study of prototypical-part networks (ProtoPNets), we create a new framework for integrating components of prototypical-part models -- ProtoPNeXt. Using ProtoPNeXt, we show that applying Bayesian hyperparameter tuning and an angular prototype similarity metric to the original ProtoPNet is sufficient to produce new state-of-the-art accuracy for prototypical-part models on CUB-200 across multiple backbones. We further deploy this framework to jointly optimize for accuracy and prototype interpretability as measured by metrics included in ProtoPNeXt. Using the same resources, this produces models with substantially superior semantics and changes in accuracy between +1.3% and -1.5%. The code and trained models will be made publicly available upon publication.