Your Location：
Home >
Browse articles >
An adaptive outlier correction quantization method for vision Transformers
Regular Papers | Updated：2025-11-18

    • An adaptive outlier correction quantization method for vision Transformers

      Enhanced Publication
    • 一种面向视觉Transformers的自适应离群值校正量化方法
    • “In the field of model compression, researchers have proposed an adaptive outlier correction quantization (AOCQ) method for Transformers, which significantly alleviates the adverse effects of outliers and reduces quantization error. This provides a solution to deploy Transformers on resource-constrained devices with minimal accuracy loss and faster runtime.”
    • Frontiers of Information Technology & Electronic Engineering   Vol. 26, Issue 10, Pages: 1879-1895(2025)

    • DOI：10.1631/FITEE.2400994    

      CLC： TP391

    • Received：11 November 2024，

      Revised：2025-07-08，

      Published：2025-10

    Scan QR Code

  • Zheyang LI, Chaoxiang LAN, Kai ZHANG, et al. An adaptive outlier correction quantization method for vision Transformers[J]. Frontiers of Information Technology & Electronic Engineering, 2025, 26(10): 1879-1895. DOI： 10.1631/FITEE.2400994.

  •  
  •  

0

Views

247

Downloads

0

CSCD

PlumX

>
Alert me when the article has been cited
Submit
Tools
Download
Export Citation
Share
Add to favorites
Add to my album

Related Articles

Transformer in reinforcement learning for decision-making: a survey
A robust tensor watermarking algorithm for diffusion-tensor images
Style-conditioned music generation with Transformer-GANs
Reversible data hiding using a transformer predictor and an adaptive embedding strategy

Related Author

Weiwei ZHAO
Peng LI
Zhenzhen HU
Dawei FENG
Shaofei CHEN
Jiaxing CHEN
Weilin YUAN
Zhi LI

Related Institution

College of Information and Communication, National University of Defense Technology
College of Intelligence Science and Technology, National University of Defense Technology
Laboratory for Parallel and Distributed Processing, National University of Defense Technology
State Key Laboratory of Public Big Data, College of Computer Science and Technology, Guizhou University
School of Electronic and Information Engineering, South China University of Technology
0