Metraipolis › archive › dc_9dff6ef123d9

COSMO: Consensus-Driven Shift Modulation for Source-Free Domain Adaptation — arXiv:2608.04604

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.04604
License (as archived)
CC0-1.0
Archived
2026-10-31T11:48:00.742Z
SHA-256 of the archived text
dd37cc2c6770fa05b3bdd86955a10d2f0d4b2f3e2a540e84d102ac1d60a84c96
Stable URL of this page
https://metraipolis.com/documento/dc_9dff6ef123d9

How to cite

COSMO: Consensus-Driven Shift Modulation for Source-Free Domain Adaptation — arXiv:2608.04604 — arXiv (en), https://arxiv.org/abs/2608.04604 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_9dff6ef123d9 (Metraipolis archive dc_9dff6ef123d9, sha256 dd37cc2c6770fa05b3bdd86955a10d2f0d4b2f3e2a540e84d102ac1d60a84c96)
BibTeX
@misc{dc9dff6ef123d9,
  title = {COSMO: Consensus-Driven Shift Modulation for Source-Free Domain Adaptation — arXiv:2608.04604},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.04604},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_9dff6ef123d9; sha256 dd37cc2c6770fa05b3bdd86955a10d2f0d4b2f3e2a540e84d102ac1d60a84c96},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.04604. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Source-free domain adaptation (SFDA) adapts a source-trained model to an unlabeled target domain without source data, a practical setting under privacy or storage constraints. Yet its self-generated supervision can reinforce source bias under substantial domain shifts. Pretrained vision-language models (VLMs) offer complementary semantic knowledge, but the relative reliability of the source model and VLM varies across target samples. Existing cross-model guidance does not explicitly account for this variation and may overwrite valid source-derived evidence under conflict, a failure we term source-derived evidence forgetting. We formulate VLM-guided SFDA as a sample-wise reliability-allocation problem and propose Consensus-Driven Shift Modulation (COSMO). COSMO replaces expert-to-expert guidance with co-adaptation through an anchored shared consensus. It first forms a sample-specific initial consensus that favors the more concentrated prediction. During adaptation, COSMO re-aggregates both branches' evolving evidence and regulates how far the resulting consensus moves from its initial anchor based on consensus uncertainty and training progress. This keeps the shared supervision anchored yet adaptive. Across four benchmarks, COSMO achieves state-of-the-art performance under matched VLM backbones. Further analyses indicate that it better balances the retention of valid source-derived evidence with the absorption of complementary VLM evidence.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-08-06 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2608.04604 — 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 2026-10-31T11:37:40.690Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-08-06?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).