Metraipolis › archive › dc_3ca979990bd1

SAGE: Salient Factor Discovery and Generation with Visual Foundation Representations — arXiv:2609.39635

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.39635
License (as archived)
CC0-1.0
Archived
2026-10-07T13:47:44.217Z
SHA-256 of the archived text
75ffc79d8460be598423d260fcc6c6093cbbcb81f1bc964d9403bb19adfe1c65
Stable URL of this page
https://metraipolis.com/documento/dc_3ca979990bd1

How to cite

SAGE: Salient Factor Discovery and Generation with Visual Foundation Representations — arXiv:2609.39635 — arXiv (en), https://arxiv.org/abs/2609.39635 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3ca979990bd1 (Metraipolis archive dc_3ca979990bd1, sha256 75ffc79d8460be598423d260fcc6c6093cbbcb81f1bc964d9403bb19adfe1c65)
BibTeX
@misc{dc3ca979990bd1,
  title = {SAGE: Salient Factor Discovery and Generation with Visual Foundation Representations — arXiv:2609.39635},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.39635},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3ca979990bd1; sha256 75ffc79d8460be598423d260fcc6c6093cbbcb81f1bc964d9403bb19adfe1c65},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.39635. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Given a target dataset, such as faces with eyeglasses, and a background dataset, such as faces without, contrastive analysis separates \textit{salient} factors specific to the target from \textit{common} content shared by both. We aim for salient representations that capture target-specific detail in each image, such as the shape, color, and position of the glasses, so that they reveal subtypes without subtype labels and guide the generation of new examples of a discovered subtype, even one with no name or text description. We introduce SAGE, which learns both factors directly in the high-dimensional spatial latent of a frozen representation autoencoder and conditions a diffusion transformer on the learned salient representation of a reference image. On Digits-ImageNet and FFHQ eyeglasses, SAGE combines high-fidelity \textit{reconstruction} (rFID below $2$) with unsupervised \textit{subtype discovery}, recovering the digits better than baselines (probe accuracy $0.950$ vs.\ at most $0.281$) and revealing eyewear types, finer sunglasses styles, and mislabeled images; salient-conditioned \textit{generation} raises Digits-ImageNet subtype accuracy over the unfactorized latent ($90.5\%$ vs.\ $27.7\%$) and diversity on both datasets. On retinal OCT, SAGE's salient space separates three diseases using only normal/disease labels.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-01 in cs.LG).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.39635 — 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 2026-10-07T13:18:10.665Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).