Importance-Aware OBS Pruning for Diffusion Models
Ba-Thinh Lam, Srijan Das, Hieu Le
Abstract
We propose importance-aware pruning for diffusion models, a training-free framework that prioritizes preserving parameters critical to semantically salient image regions. To do so, we incorporate spatial importance maps -- derived from conditioning signals or model attention -- into the pruning objective. This produces parameter rankings aligned with perceptual relevance rather than uniform reconstruction error. On MS-COCO dataset, our proposed approach consistently retains subject fidelity and structural correctness at high compression ratios where conventional pruning causes visible degradation. These results demonstrate that content-aware objectives are key to perceptually faithful compression of generative models.
Topics
Classified with taxonomy v2 on Wed, 2 Sept 2026.