Metraipolis › archive › dc_0bc4d8160aca
LCM-UNet: Scale-Selective Mamba with Reparameterizable Multi-Basis Local Modeling for Lightweight Skin Lesion Segmentation — arXiv:2611.04409
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.04409
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-04T19:03:13.736Z
- SHA-256 of the archived text
12e3c2b5512a52ed53e346de6a51817ffeebddaed39c3024f925ca3a3bb0aabb- Stable URL of this page
https://metraipolis.com/documento/dc_0bc4d8160aca
How to cite
LCM-UNet: Scale-Selective Mamba with Reparameterizable Multi-Basis Local Modeling for Lightweight Skin Lesion Segmentation — arXiv:2611.04409 — arXiv (en), https://arxiv.org/abs/2611.04409 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0bc4d8160aca (Metraipolis archive dc_0bc4d8160aca, sha256 12e3c2b5512a52ed53e346de6a51817ffeebddaed39c3024f925ca3a3bb0aabb)
BibTeX
@misc{dc0bc4d8160aca,
title = {LCM-UNet: Scale-Selective Mamba with Reparameterizable Multi-Basis Local Modeling for Lightweight Skin Lesion Segmentation — arXiv:2611.04409},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.04409},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0bc4d8160aca; sha256 12e3c2b5512a52ed53e346de6a51817ffeebddaed39c3024f925ca3a3bb0aabb},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.04409. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Accurate lesion segmentation underpins automated dermoscopic analysis: masks enable measurements of asymmetry, border and size. Deployment in primary care or on mobile devices requires a compact network. Existing ultra-light Mamba U-Nets offer global context, but global-modeling cost depends on each Mamba block's resolution, while their local branches rely on a single square kernel. We introduce LCM-UNet, a 33,474-parameter U-shaped network that fuses every skip at its native resolution, restricts bidirectional Mamba to stages at one-sixteenth of the input resolution or coarser, and uses local convolutions at finer scales. A multi-basis reparameterizable adapter augments each local branch during training with dilated, horizontal and vertical depthwise kernels, then folds them into the host kernel before inference. Each skip connection also learns a stage-specific gain. Nine networks were retrained on the official ISIC 2017 and ISIC 2018 partitions under a harmonized protocol, with PH$^2$ reserved for external validation and ablations run on a development split. LCM-UNet achieved Dice scores of 88.62%, 89.73% and 90.46% on ISIC 2017, ISIC 2018 and PH$^2$, respectively, ranking first on both ISIC test sets and second on PH$^2$ among the compared models. In the native-resolution host, removing Mamba from the one-eighth-resolution decoder stage eliminated 63% of Mamba operations, corresponding to 5.3% of all operations, with a Dice change of $-0.03$ percentage points. The adapter improved Dice by 0.48 percentage points over a single host kernel without adding deployed parameters or operations. With latency so far measured only on a GPU, LCM-UNet is a candidate for resource-constrained deployment. Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2611.04409 — 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 2026-11-04T18:39:37.501Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).