S $^{2}$ Q-VDiT$^+$: Accurate Quantized Video Diffusion Transformer with Multi-Resolution Sampling and Structural Distillation.

Authors

Weilun Feng, Chuanguang Yang, Haotong Qin, Yuqi Li, Mingqiang Wu, Guoxin Fan, Xiangqi Li, Zhulin An, Libo Huang, Boyu Diao, Renshuai Tao, Fuzhen Zhuang, Michele Magno, Yongjun Xu, Zhigang Zeng, Tingwen Huang, Yingli Tian

Published in

IEEE transactions on pattern analysis and machine intelligence. Volume PP. Aug 11, 2026. Epub Aug 11, 2026.

Abstract

Large-scale video diffusion models (V-DMs) have achieved remarkable text-to-video generation quality, yet their massive computational complexity makes deployment costly. Post-Training Quantization (PTQ) offers an appealing route to accelerate inference without retraining, but existing diffusion PTQ methods remain fragile on modern V-DMs. A key reason is that contemporary V-DMs are intrinsically multi-resolution due to multi-stage training, while most prior PTQ pipelines calibrate at a fixed resolution, causing suboptimal calibration signals and biased distributions under resolution changes. To address this gap, we propose S $^{2}$ Q-VDiT $^+$, a multi-resolution co-design PTQ framework from data, supervision, and quantizer perspectives. First, Denoising-Prior Based Multi-Resolution Sampling constructs resolution-consistent noisy latents by mapping to the clean space and re-noising, together with a trajectory-aware resolution policy across timesteps. Second, Structure-Aware Multi-Resolution Distillation enhances structural alignment via window-wise distillation and transfers resolution-aware spatial dependencies via multi-scale attention distillation. Third, Debiased Modulated Quantization mitigates skewed distributions using asymmetric weight quantization and a fuseable activation debiasing scheme. Extensive experiments on multiple state-of-the-art video generation models demonstrate that S$^{2}$ Q-VDiT$^+$ consistently outperforms strong PTQ baselines under W4A6 and W4A4, delivers up to $2.08\times$ end-to-end speedup, and reduces model storage and inference memory by up to $3.8\times$ and $2.1\times$, respectively.

PMID:
42579580
Bibliographic data and abstract were imported from PubMed on 12 Aug 2026.

Read full publication at:
Please sign in to see all details.

Stats

1-terrible, 9-excellent. How would you rate this publication? Sign in in to submit your rating.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.