PoMQ-ViT: Mixed-Precision Quantization Vision Transformer with Pareto Optimization
Applied Sciences · 09.09.2025 · 15 (18) · 9856
DOI: 10.3390/app15189856 ↗Kurz gefasst
Fachartikel in der Zeitschrift Applied Sciences, online erschienen am 09.09.2025. Verfasst von Zhiqiang Wu und 1 weiteren Autor:innen. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 41 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Vision Transformer (ViT) has established itself as the leading model in image processing, and quantization offers a practical way to deploy ViTs on memory-constrained devices. This technique maps float32 inference to lower bits such as int8. However, existing studies often use uniform bit-width across ViT encoders, leading to sub-optimal allocation and accuracy loss. This study aims to optimize bit-width allocation by leveraging the observed variation in quantization sensitivity among different ViT encoders. We propose PoMQ-ViT, a method that dynamically assigns bit-width to ViT encoders based on their sensitivity. It uses Pareto Optimization to balance accuracy and inference speed. Experiments on the ImageNet Dataset show PoMQ-ViT outperforms uniform bit-width methods across DeiT and Swin models. Within the same computation budget, it also achieves 0.1–1.5 percentage points higher accuracy than other mixed-precision paradigms across different models. This work demonstrates PoMQ-ViT’s effectiveness in ViT quantization, providing a practical solution for resource-constrained deployments.
Zitieren
APA
Wu, Z., & Zhao, Z. (2025). PoMQ-ViT: Mixed-Precision Quantization Vision Transformer with Pareto Optimization. Applied Sciences, 15(18), 9856. https://doi.org/10.3390/app15189856
Vancouver
Wu Z, Zhao Z. PoMQ-ViT: Mixed-Precision Quantization Vision Transformer with Pareto Optimization. Applied Sciences. 2025;15(18):9856. doi:10.3390/app15189856
BibTeX
@article{Wu2025,
author = {Wu, Zhiqiang and Zhao, Zhong},
title = {PoMQ-ViT: Mixed-Precision Quantization Vision Transformer with Pareto Optimization},
journal = {Applied Sciences},
year = {2025},
volume = {15},
number = {18},
pages = {9856},
doi = {10.3390/app15189856}
}
EUVIMED