Metraipolis › archive › dc_2e4fc0b75dbf
Generative Uncertainty as a Self-supervised Signal for Semantic Similarity Learning — arXiv:2609.35341
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.35341
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T23:52:42.328Z
- SHA-256 of the archived text
dff2c26cb86bd391f96fb9a03de9f49b1fec26de67a65449343ee39c5b8c8a26- Stable URL of this page
https://metraipolis.com/documento/dc_2e4fc0b75dbf
How to cite
Generative Uncertainty as a Self-supervised Signal for Semantic Similarity Learning — arXiv:2609.35341 — arXiv (en), https://arxiv.org/abs/2609.35341 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2e4fc0b75dbf (Metraipolis archive dc_2e4fc0b75dbf, sha256 dff2c26cb86bd391f96fb9a03de9f49b1fec26de67a65449343ee39c5b8c8a26)
BibTeX
@misc{dc2e4fc0b75dbf,
title = {Generative Uncertainty as a Self-supervised Signal for Semantic Similarity Learning — arXiv:2609.35341},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.35341},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2e4fc0b75dbf; sha256 dff2c26cb86bd391f96fb9a03de9f49b1fec26de67a65449343ee39c5b8c8a26},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.35341. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Evaluating semantic similarity between videos is a fundamental challenge in computer vision, essential for tasks ranging from out-of-distribution (OOD) detection to video retrieval. However, defining and labeling video similarity is notoriously difficult and expensive due to the complex spatio-temporal nature. In this paper, we propose a novel self-supervised approach that leverages generative uncertainty from text-to-video (T2V) diffusion models to learn semantic similarity without human annotations. Our method is based on the observation that T2V models produce consistent outputs for familiar concepts but exhibit high variance and uncertainty when prompted with specialized concepts. We utilize this behavior to identify stable semantic features within existing pretrained representations, such as VideoMAE and V-JEPA. Specifically, we learn a mask over these embeddings using purely generated data, encouraging the model to retain features that remain consistent across generations of general concepts while discarding those associated with generative noise or uncertainty. Experimental results across three key tasks demonstrate that our learned feature subspaces consistently outperform original pretrained features and baseline feature selection methods. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-29 in cs.CV). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2609.35341 — 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 2026-10-07T23:29:27.554Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).