Szczegóły publikacji
Opis bibliograficzny
A study on highly efficient compact transformer features for histopathological image recognition / Didih Rizki CHANDRANEGARA, Przemysław NIEDZIELA, Bogusław CYGANEK // W: Artificial Intelligence and Soft Computing : 24th International Conference, ICAISC 2025 : Zakopane, Poland, June 22–26, 2025 : proceedings , Pt. 2 / eds. Leszek Rutkowski, [et al.]. — Cham : Springer Nature Switzerland, cop. 2026. — ( Lecture Notes in Computer Science ; ISSN 0302-9743. Lecture Notes in Artificial Intelligence ; 15949 ). — ISBN: 978-3-032-03707-7; e-ISBN: 978-3-032-03708-4. — S. 38–49. — Bibliogr., Abstr. — Publikacja dostępna online od: 2025-11-01
Autorzy (3)
Słowa kluczowe
Dane bibliometryczne
| ID BaDAP | 164440 |
|---|---|
| Data dodania do BaDAP | 2026-01-22 |
| DOI | 10.1007/978-3-032-03708-4_3 |
| Rok publikacji | 2026 |
| Typ publikacji | materiały konferencyjne (aut.) |
| Otwarty dostęp |  |
| Wydawca | Springer |
| Konferencja | International Conference on Artificial Intelligence and Soft Computing 2025 |
| Czasopismo/seria | Lecture Notes in Computer Science |
Abstract
Histopathological image recognition requires efficient feature extraction and dimensionality reduction to manage the complexity of the scans. For this, we propose a hybrid models that combine TinyViT with the self-supervised learning capabilities of DINOv1 and dimensionality reduction techniques, which achieve higher accuracy with improved computational efficiency outperforming DINOv1 architectures. To enhance performance, feature reduction techniques, such as PCA and NCA, are employed to reduce feature sizes while minimizing accuracy loss and retaining critical information. Although DINOv1 exhibits state-of-the-art accuracy in general computer vision tasks, its performance on medical images at the researched magnification is limited. In contrast, TinyVIT-based models offer a balanced solution to efficiently process large histopathology scans with improved accuracy and reduced computational requirements, as our results show.
