ProtoS-ViT: Visual foundation models for sparse self-explainable classifications
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866912155474853888 |
|---|---|
| author | Turbé, Hugues Bjelogrlic, Mina Mengaldo, Gianmarco Lovis, Christian |
| author_facet | Turbé, Hugues Bjelogrlic, Mina Mengaldo, Gianmarco Lovis, Christian |
| contents | Prototypical networks aim to build intrinsically explainable models based on the linear summation of concepts. Concepts are coherent entities that we, as humans, can recognize and associate with a certain object or entity. However, important challenges remain in the fair evaluation of explanation quality provided by these models. This work first proposes an extensive set of quantitative and qualitative metrics which allow to identify drawbacks in current prototypical networks. It then introduces a novel architecture which provides compact explanations, outperforming current prototypical models in terms of explanation quality. Overall, the proposed architecture demonstrates how frozen pre-trained ViT backbones can be effectively turned into prototypical models for both general and domain-specific tasks, in our case biomedical image classifiers. Code is available at \url{https://github.com/hturbe/protosvit}. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_10025 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | ProtoS-ViT: Visual foundation models for sparse self-explainable classifications Turbé, Hugues Bjelogrlic, Mina Mengaldo, Gianmarco Lovis, Christian Computer Vision and Pattern Recognition Machine Learning Prototypical networks aim to build intrinsically explainable models based on the linear summation of concepts. Concepts are coherent entities that we, as humans, can recognize and associate with a certain object or entity. However, important challenges remain in the fair evaluation of explanation quality provided by these models. This work first proposes an extensive set of quantitative and qualitative metrics which allow to identify drawbacks in current prototypical networks. It then introduces a novel architecture which provides compact explanations, outperforming current prototypical models in terms of explanation quality. Overall, the proposed architecture demonstrates how frozen pre-trained ViT backbones can be effectively turned into prototypical models for both general and domain-specific tasks, in our case biomedical image classifiers. Code is available at \url{https://github.com/hturbe/protosvit}. |
| title | ProtoS-ViT: Visual foundation models for sparse self-explainable classifications |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2406.10025 |