Metraipolis › archive › dc_4ec4fc7b9522
SAKURA: Serving Attention Key-Value State Under Rate Adaptation for Private Image Editing on Phones — arXiv:2610.29733
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.29733
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-26T19:54:39.453Z
- SHA-256 of the archived text
923d7f7e87f695be8b80b3d0c4dd3df94684c796e489dbb475bbcb8dc4695ff4- Stable URL of this page
https://metraipolis.com/documento/dc_4ec4fc7b9522
How to cite
SAKURA: Serving Attention Key-Value State Under Rate Adaptation for Private Image Editing on Phones — arXiv:2610.29733 — arXiv (en), https://arxiv.org/abs/2610.29733 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_4ec4fc7b9522 (Metraipolis archive dc_4ec4fc7b9522, sha256 923d7f7e87f695be8b80b3d0c4dd3df94684c796e489dbb475bbcb8dc4695ff4)
BibTeX
@misc{dc4ec4fc7b9522,
title = {SAKURA: Serving Attention Key-Value State Under Rate Adaptation for Private Image Editing on Phones — arXiv:2610.29733},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.29733},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_4ec4fc7b9522; sha256 923d7f7e87f695be8b80b3d0c4dd3df94684c796e489dbb475bbcb8dc4695ff4},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.29733. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Diffusion transformers are widely used in image-editing services. To preserve both efficiency and privacy, existing systems commonly rely on cloud-device collaboration. Keeping request-specific inference on the device avoids sending personal images and prompts to the cloud. These workloads often use common templates, such as in face swapping and virtual try-on, enabling KV state to be precomputed and reused across requests. However, this reusable state is large and scales with the number of templates, making it impractical to store all of them on mobile devices. Storing them in the cloud addresses this capacity issue, but cache transfer is placed onto the inference path and can even be slower than on-device computation, especially under fluctuating mobile networks. We introduce SAKURA, a cloud-edge architecture that delivers reusable attention state to mobile devices under a bandwidth budget. It first uses a fine-grained token-importance-aware compression method to preserve image quality under limited transfer budgets. At runtime, SAKURA adaptively selects the cache representation for each upcoming shard based on network bandwidth. To enable efficient execution, we develop a pipelined runtime that overlaps cache transfer and reconstruction with on-device execution. We evaluate SAKURA using three editing tasks and two diffusion models, using a Snapdragon 8 Elite Gen 5 phone. SAKURA reduces latency by 1.82x and 3.27x over CacheGen adapted to DiT and on-device generation, respectively. It also reduces cache storage by 3.06x over uncompressed cache, while preserving image quality. Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.DC, nuovi lavori del giorno). Categorie arXiv: Software Engineering (cs.SE) ; Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC); Networking and Internet Architecture (cs.NI). Pagina del lavoro: https://arxiv.org/abs/2610.29733 — 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 2026-10-26T19:46:03.299Z dalla pagina https://arxiv.org/list/cs.DC/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).