Contour wavelet diffusion - a fast and high-quality facial expression generation model

被引:0
|
作者
Xu, Chenwei [1 ]
Zou, Yuntao [2 ,3 ]
机构
[1] Commun Univ Zhejiang, Sch Design & Art, Hangzhou, Peoples R China
[2] Huazhong Univ Sci & Technol, Sch Comp Sci & Technol, Wuhan, Peoples R China
[3] Huazhong Univ Sci & Technol, Sch Comp Sci & Technol, Luoyu Rd 1037, Wuhan 430074, Peoples R China
关键词
Facial Expression Generation; diffusion model; contour wavelet; TRANSFORM; DESIGN;
D O I
10.1080/09540091.2024.2316023
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Facial expressions are important for conveying information in human interactions. The diffusion model can generate high-quality images for clearer and more discriminative faces, but its training and inference time is often prolonged, hampering practical application. Latent space diffusion models have shown promise in speeding up training by leveraging feature space parameters, but they require additional network structures. To address these limitations, we propose a contour wavelet diffusion model that accelerates both training and inference speeds. We use a contour wavelet transform to extract components from images and features, achieving substantial acceleration while preserving reconstruction quality. A normalised random channel attention module enhances the quality of generated images by focusing on high-frequency information. We also include a reconstruction loss function to enhance convergence speed. Experimental results demonstrate the effectiveness of our approach in boosting the training and inference speeds of diffusion models without sacrificing image quality. Fast generation of facial expressions can provide a smoother and more natural user experience, which is important for real-time applications. In addition, the increase in inference speed can save the use of computational resources, reduce system cost and improve energy efficiency, which is conducive to promoting the development and application of this technology.
引用
收藏
页数:20
相关论文
共 50 条
  • [1] Contour wavelet diffusion: A fast and high-quality image generation model
    Ding, Yaoyao
    Zhu, Xiaoxi
    Zou, Yuntao
    [J]. COMPUTATIONAL INTELLIGENCE, 2024, 40 (02)
  • [2] High-quality facial-expression image generation for UAV pedestrian detection
    Tang, Yumin
    Fan, Jing
    Qu, Jinshuai
    [J]. FRONTIERS IN SPACE TECHNOLOGIES, 2022, 3
  • [3] Efficient image generation with Contour Wavelet Diffusion
    Zhang, Dimeng
    Li, JiaYao
    Chen, Zilong
    Zou, Yuntao
    [J]. Computers and Graphics (Pergamon), 2024, 124
  • [4] ProDiff: Progressive Fast Diffusion Model for High-Quality Text-to-Speech
    Huang, Rongjie
    Zhao, Zhou
    Liu, Huadai
    Liu, Jinglin
    Cui, Chenye
    Ren, Yi
    [J]. PROCEEDINGS OF THE 30TH ACM INTERNATIONAL CONFERENCE ON MULTIMEDIA, MM 2022, 2022, : 2595 - 2605
  • [5] A fast and high-quality charge model for the next generation general AMBER force field
    He, Xibing
    Man, Viet H.
    Yang, Wei
    Lee, Tai-Sung
    Wang, Junmei
    [J]. JOURNAL OF CHEMICAL PHYSICS, 2020, 153 (11):
  • [6] VideoFusion: Decomposed Diffusion Models for High-Quality Video Generation
    Luo, Zhengxiong
    Chen, Dayou
    Zhang, Yingya
    Huang, Yan
    Wang, Liang
    Shen, Yujun
    Zhao, Deli
    Zhou, Jingren
    Tan, Tieniu
    [J]. 2023 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2023, : 10209 - 10218
  • [7] scDiffusion: conditional generation of high-quality single-cell data using diffusion model
    Luo, Erpai
    Hao, Minsheng
    Wei, Lei
    Zhang, Xuegong
    [J]. BIOINFORMATICS, 2024, 40 (09)
  • [8] Efficient, High-Quality Image Contour Detection
    Catanzaro, Bryan
    Su, Bor-Yiing
    Sundaram, Narayanan
    Lee, Yunsup
    Murphy, Mark
    Keutzer, Kurt
    [J]. 2009 IEEE 12TH INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2009, : 2381 - 2388
  • [9] FAST REACTIONS AND HIGH-QUALITY
    DUNN, J
    [J]. ENGINEER, 1984, 258 (6690): : 69 - 70
  • [10] Fast High-Quality Noise
    Frisvad, Jeppe Revall
    Wyvill, Geoff
    [J]. GRAPHITE 2007: 5TH INTERNATIONAL CONFERENCE ON COMPUTER GRAPHICS AND INTERACTIVE TECHNIQUES IN AUSTRALASIA AND SOUTHERN ASIA, PROCEEDINGS, 2007, : 243 - +