EUVIMEDEUVIMED entdecken
Crossref

Accelerating Controllable Generation via Hybrid-grained Cache

Lin Liu, Huixia Ben, Shuo Wang, Jinda Lu, Junxiang Qiu, Shengeng Tang, Yanbin Hao

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (9) · 7215-7223

DOI: 10.1609/aaai.v40i9.37658 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Lin Liu und 6 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Controllable generative models have been widely used to improve the realism of synthetic visual content. However, such models must handle control conditions and content generation computational requirements, resulting in generally low generation efficiency. To address this issue, we propose a Hybrid-Grained Cache (HGC) approach that reduces computational overhead by adopting cache strategies with different granularities at different computational stages. Specifically, (1) we use a coarse-grained cache (block-level) based on feature reuse to dynamically bypass redundant computations in encoder-decoder blocks between each step of model reasoning. (2) We design a fine-grained cache (prompt-level) that acts within a module, where the fine-grained cache reuses cross-attention maps within consecutive reasoning steps and extends them to the corresponding module computations of adjacent steps. These caches of different granularities can be seamlessly integrated into each computational link of the controllable generation process. We verify the effectiveness of HGC on four benchmark datasets, especially its advantages in balancing generation efficiency and visual quality. For example, on the COCO-Stuff segmentation benchmark, our HGC significantly reduces the computational cost (MACs) by 63% (from 18.22T → 6.70T↓), while keeping the loss of semantic fidelity (quantized performance degradation) within 1.5%.

Zitieren

APA

Liu, L., Ben, H., Wang, S., Lu, J., Qiu, J., Tang, S., & Hao, Y. (2026). Accelerating Controllable Generation via Hybrid-grained Cache. Proceedings of the AAAI Conference on Artificial Intelligence, 40(9), 7215-7223. https://doi.org/10.1609/aaai.v40i9.37658

Vancouver

Liu L, Ben H, Wang S, Lu J, Qiu J, Tang S, et al. Accelerating Controllable Generation via Hybrid-grained Cache. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(9):7215-7223. doi:10.1609/aaai.v40i9.37658

BibTeX
@article{Liu2026,
  author = {Liu, Lin and Ben, Huixia and Wang, Shuo and Lu, Jinda and Qiu, Junxiang and Tang, Shengeng and Hao, Yanbin},
  title = {Accelerating Controllable Generation via Hybrid-grained Cache},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {9},
  pages = {7215--7223},
  doi = {10.1609/aaai.v40i9.37658}
}
Kostenlos starten →