Metraipolis › archive › dc_95d1c8a5fc82
PoE-Fuse: Precision-Weighted Expert Fusion for Bi-Temporal Change Understanding — arXiv:2609.37485
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.37485
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T17:32:56.700Z
- SHA-256 of the archived text
76f9c4dde93b5143c3f88022e719283531815d05831cf9c64a93751468ff2372- Stable URL of this page
https://metraipolis.com/documento/dc_95d1c8a5fc82
How to cite
PoE-Fuse: Precision-Weighted Expert Fusion for Bi-Temporal Change Understanding — arXiv:2609.37485 — arXiv (en), https://arxiv.org/abs/2609.37485 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_95d1c8a5fc82 (Metraipolis archive dc_95d1c8a5fc82, sha256 76f9c4dde93b5143c3f88022e719283531815d05831cf9c64a93751468ff2372)
BibTeX
@misc{dc95d1c8a5fc82,
title = {PoE-Fuse: Precision-Weighted Expert Fusion for Bi-Temporal Change Understanding — arXiv:2609.37485},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.37485},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_95d1c8a5fc82; sha256 76f9c4dde93b5143c3f88022e719283531815d05831cf9c64a93751468ff2372},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.37485. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Bi-temporal change understanding, which localizes and characterizes what changed between two satellite images, is central to disaster response and environmental monitoring, spanning change detection, building localization, and damage assessment. Strong vision-language models address these tasks, but adapting them typically requires full fine-tuning or reinforcement learning, which is costly and unstable. We propose PoE-Fuse, a parameter-efficient framework that instead composes frozen foundation experts for geometry, grounding, and language, resampling their features onto a shared spatial grid and training only a lightweight fusion trunk. PoE-Fuse treats the aligned features as Gaussian observations of a latent scene state and fuses them by learned per-cell precision. This product-of-experts estimator strictly generalizes uniform summation and scalar gating, and extends to change fields by composing the precisions of the two timestamps. A single shared trunk solves the three tasks at once, reaching a mean F1 of 59.2%, compared with 40.7% for an instruction-tuned temporal vision-language assistant, and surpassing dedicated change-detection models retrained under the same protocol and training budget. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.LG). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.37485 — 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 2026-10-07T16:59:48.552Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).