LinCa: Accelerating Diffusion Models via Learnable Decomposed Feature Caching
초록
Diffusion models have achieved remarkable success in image and video generation, yet the high computational cost of iterative sampling remains a critical bottleneck for practical deployment. Feature caching has emerged as a promising acceleration paradigm by reusing or predicting intermediate features across timesteps. However, existing training-free methods apply uniform prediction strategies that cannot adapt to the heterogeneous feature dynamics, causing significant quality degradation under high acceleration ratios. We propose LinCa, a feature caching framework based on learnable invertible networks. LinCa decomposes cached features into sub-components with distinct continuity properties via a lightweight invertible network and applies differentiated prediction orders matched to each component. The strict invertibility guarantees lossless reconstruction back to the original feature space, forming a unified Decompose-Predict-Reconstruct pipeline. By training separate predictors for different models and timestep segments, LinCa adapts to heterogeneous feature dynamics. Experiments on FLUX, Qwen-Image, and HunyuanVideo demonstrate that LinCa, with less than 0.2% additional parameters, significantly outperforms existing methods and maintains near-lossless quality at 5-7x speedup. Code: https://github.com/QHR69/LinCa
저자 (12명)
- Jinshan Liu — LinkedIn 검색
- Haoran Qin — LinkedIn 검색
- Xiaobing Tu — LinkedIn 검색
- Jiacheng Liu — LinkedIn 검색
- Jiahui Hu — LinkedIn 검색
- Zhengan Yan — LinkedIn 검색
- Yukun Xie — LinkedIn 검색
- Kerui Shen — LinkedIn 검색
- Jinkui Ren — LinkedIn 검색
- Yuqi Lin — LinkedIn 검색
- Xiantao Zhang — LinkedIn 검색
- Linfeng Zhang — LinkedIn 검색
저자 LinkedIn 변경 추적은 추후 자동화 예정입니다. 현재는 검색 링크를 제공합니다.