Metraipolis › archive › dc_08176d7286cd

Ultra-Lightweight 3D Medical Image Segmentation via Knowledge Distillation in Label-Scarce Domains — arXiv:2611.10033

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.10033
License (as archived)
CC0-1.0
Archived
2026-11-10T18:42:11.862Z
SHA-256 of the archived text
708e37c19db24941caae5f928e1f646da922a7323d5ce4e9e42804f6d1f1a78b
Stable URL of this page
https://metraipolis.com/documento/dc_08176d7286cd

How to cite

Ultra-Lightweight 3D Medical Image Segmentation via Knowledge Distillation in Label-Scarce Domains — arXiv:2611.10033 — arXiv (en), https://arxiv.org/abs/2611.10033 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_08176d7286cd (Metraipolis archive dc_08176d7286cd, sha256 708e37c19db24941caae5f928e1f646da922a7323d5ce4e9e42804f6d1f1a78b)
BibTeX
@misc{dc08176d7286cd,
  title = {Ultra-Lightweight 3D Medical Image Segmentation via Knowledge Distillation in Label-Scarce Domains — arXiv:2611.10033},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.10033},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_08176d7286cd; sha256 708e37c19db24941caae5f928e1f646da922a7323d5ce4e9e42804f6d1f1a78b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.10033. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Mamba is a neural network architecture that reduces computational complexity, and has been shown to be effective for 3D medical image segmentation. However, existing models still remain computationally expensive. We propose compressing Mamba by distillation while achieving competitive performance. Specifically, we demonstrate compression of a 139M-parameter SegMamba-V2 teacher into a lightweight TinyUNet3D student with about 0.96M parameters, achieving about 145$\times$ compression with a performance drop of just over one Dice point. We further investigate the utility of knowledge distillation in a range of data-labeling scenarios under identical experimental settings, and show that its advantage is most prominent when labeled data is scarce. To demonstrate the generalizability of our approach, experiments are conducted on three datasets: BraTS 2023, MSD Spleen and MSD Hippocampus. The code for TinyUNet3D is publicly available at: this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2611.10033 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-10T18:37:37.570Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).