Metraipolis › archive › dc_eecd6535d6b0

Neural Disjunctive Normal Forms for Learning Exact Rank Filters — arXiv:2610.32203

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.32203
License (as archived)
CC0-1.0
Archived
2026-10-26T20:14:13.289Z
SHA-256 of the archived text
ee90b6429d17907f0f562484b32362986a3d1ca4f32e5291405473ed826077d1
Stable URL of this page
https://metraipolis.com/documento/dc_eecd6535d6b0

How to cite

Neural Disjunctive Normal Forms for Learning Exact Rank Filters — arXiv:2610.32203 — arXiv (en), https://arxiv.org/abs/2610.32203 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_eecd6535d6b0 (Metraipolis archive dc_eecd6535d6b0, sha256 ee90b6429d17907f0f562484b32362986a3d1ca4f32e5291405473ed826077d1)
BibTeX
@misc{dceecd6535d6b0,
  title = {Neural Disjunctive Normal Forms for Learning Exact Rank Filters — arXiv:2610.32203},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.32203},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_eecd6535d6b0; sha256 ee90b6429d17907f0f562484b32362986a3d1ca4f32e5291405473ed826077d1},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.32203. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Gradient-based learning of morphological operators is hindered by locally sparse derivatives with respect to the structuring elements. We develop a method for learning rank filters that retains exact operator evaluation while distributing the derivatives across the window. Using threshold decomposition, we construct a neural disjunctive normal form representation and establish a sufficient condition for exactness with integer-valued images and structuring elements. We then derive a smooth surrogate from the unique root of a smooth threshold equation. Implicit differentiation gives closed-form derivatives with normalised local weights, whose distribution is controlled by a smoothing parameter. These derivatives are used in a straight-through procedure: the forward pass evaluates the exact rank filter, while the backward pass uses the surrogate. For fixed ranks, the procedure learns real-valued structuring elements of any rank filter, including the median, and extends to compositions, infima, and suprema of rank filters. Experiments on image denoising using MNIST and Fashion-MNIST and retinal vessel segmentation show that, starting from a generic random initialisation, the proposed approach consistently improves performance over training with exact subgradients in every setting. In image denoising, our best network also outperforms smooth morphological networks in all settings. On vessel segmentation, the learned erosions recover structuring elements that fit the vessels, as theoretically expected, whereas most of those trained with exact subgradients remain close to their initialisation.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.32203 — 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 2026-10-26T19:49:03.471Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).