Metraipolis › archive › dc_d46a2ee10409
Task-State Adaptation with Prototype Memory for Multi-Task Dense Prediction — arXiv:2608.28078
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.28078
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T11:33:08.331Z
- SHA-256 of the archived text
ce213a06746ffbc10cee876dd696aa71a67de384f5fc1671e0a4be21b69fa882- Stable URL of this page
https://metraipolis.com/documento/dc_d46a2ee10409
How to cite
Task-State Adaptation with Prototype Memory for Multi-Task Dense Prediction — arXiv:2608.28078 — arXiv (en), https://arxiv.org/abs/2608.28078 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d46a2ee10409 (Metraipolis archive dc_d46a2ee10409, sha256 ce213a06746ffbc10cee876dd696aa71a67de384f5fc1671e0a4be21b69fa882)
BibTeX
@misc{dcd46a2ee10409,
title = {Task-State Adaptation with Prototype Memory for Multi-Task Dense Prediction — arXiv:2608.28078},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.28078},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d46a2ee10409; sha256 ce213a06746ffbc10cee876dd696aa71a67de384f5fc1671e0a4be21b69fa882},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.28078. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Vision foundation backbones provide strong representations for dense prediction, yet a single shared feature still needs to support tasks with different, image-dependent adaptation requirements. We propose MemMTL, a multi-task dense prediction framework that estimates a compact task state from global visual context and refines it through a learnable task-state prototype memory. The refined state is converted into task-conditioned expert logits and combined with token-level logits before sparse top-$k$ selection over a local expert bank shared by all tasks. A separate task-agnostic residual bank provides a common adaptation path, and both paths are added once to the backbone feature before task-specific prediction. We specify a matched evaluation protocol on NYUD-v2 and PASCAL-Context with SAM 3 and ViT-L backbones to measure predictive quality, computational cost, and the contributions of task-state conditioning, prototype retrieval, and sparse routing. The numerical record in the present working draft predates this canonical implementation and must be regenerated before it can support empirical claims. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-08-31 in cs.CV). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2608.28078 — 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 2026-10-17T11:10:05.299Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-08-31?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).