| Ayoobi, Hamed, Potyka, Nico and Toni, Francesca 2025. ProtoArgNet: Interpretable image classification with super-prototypes and argumentation. Presented at: The 39th Annual AAAI Conference on Artificial Intelligence, Philadelphia, Pennsylvania, USA, 25 February – 4 March 2025. Published in: Walsh, Toby, Shah, Julie and Kolter, Zico eds. Proceedings of the AAAI Conference on Artificial Intelligence. , vol.39 (2) Washington, DC, USA: AAAI Press, pp. 1791-1799. 10.1609/aaai.v39i2.32173 |
Preview |
PDF
- Published Version
Download (1MB) | Preview |
Abstract
We propose ProtoArgNet, a novel interpretable deep neural architecture for image classification in the spirit of prototypical-part-learning as found, e.g., in ProtoPNet. While earlier approaches associate every class with multiple prototypical-parts, ProtoArgNet uses super-prototypes that combine prototypical-parts into a unified class representation. This is done by combining local activations of prototypes in an MLP-like manner, enabling the localization of prototypes and learning (non-linear) spatial relationships among them. By leveraging a form of argumentation, ProtoArgNet is capable of providing both supporting (i.e. `this looks like that') and attacking (i.e. `this differs from that') explanations. We demonstrate on several datasets that ProtoArgNet outperforms state-of-the-art prototypical-part-learning approaches. Moreover, the argumentation component in ProtoArgNet is customisable to the user's cognitive requirements by a process of sparsification, which leads to more compact explanations compared to state-of-the-art approaches.
| Item Type: | Conference or Workshop Item (Paper) |
|---|---|
| Date Type: | Published Online |
| Status: | Published |
| Schools: | Schools > Computer Science & Informatics |
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science |
| Publisher: | AAAI Press |
| ISBN: | 978-1-57735-897-8 |
| Date of First Compliant Deposit: | 10 June 2025 |
| Last Modified: | 10 Jul 2025 01:30 |
| URI: | https://orca.cardiff.ac.uk/id/eprint/178818 |
Actions (repository staff only)
 |
Edit Item |
Dimensions
Dimensions