《计算机应用》唯一官方网站 ›› 2025, Vol. 45 ›› Issue (8): 2720-2726.DOI: 10.11772/j.issn.1001-9081.2024071039

• 多媒体计算与计算机仿真 • 上一篇    

基于空间变换网络和特征分布校准的小样本皮肤图像分类模型

王静, 刘嘉星(), 宋婉莹, 薛嘉兴, 丁温欣   

  1. 西安科技大学 通信与信息工程学院,西安 710600
  • 收稿日期:2024-07-23 修回日期:2024-10-12 接受日期:2024-10-12 发布日期:2024-11-19 出版日期:2025-08-10
  • 通讯作者: 刘嘉星
  • 作者简介:王静(1986—),女,河南安阳人,讲师,博士,CCF会员,主要研究方向:计算机视觉、雷达信号处理
    宋婉莹(1988—),女,山东聊城人,副教授,博士,主要研究方向:图像处理、计算机视觉
    薛嘉兴(2000—),男,陕西宝鸡人,硕士研究生,主要研究方向:图像处理、计算机视觉
    丁温欣(2002—),女,陕西渭南人,硕士研究生,主要研究方向:计算机视觉。
  • 基金资助:
    国家自然科学基金资助项目(61901358)

Few-shot skin image classification model based on spatial transformer network and feature distribution calibration

Jing WANG, Jiaxing LIU(), Wanying SONG, Jiaxing XUE, Wenxin DING   

  1. College of Communication and Information Engineering,Xi’an University of Science and Technology,Xi’an Shaanxi 710600,China
  • Received:2024-07-23 Revised:2024-10-12 Accepted:2024-10-12 Online:2024-11-19 Published:2025-08-10
  • Contact: Jiaxing LIU
  • About author:WANG Jing, born in 1986, Ph. D., lecturer. Her research interests include computer vision, radar signal processing.
    SONG Wanying, born in 1988, Ph. D., associate professor. Her research interests include image processing, computer vision.
    XUE Jiaxing, born in 2000, M. S. candidate. His research interests include image processing, computer vision.
    DING Wenxin, born in 2002, M. S. candidate. Her research interests include computer vision.
  • Supported by:
    National Natural Science Foundation of China(61901358)

摘要:

基于深度学习的图像分类模型通常需要大量标记数据,然而,在医学领域的皮肤病变分类任务中,收集大量图像数据面临着诸多挑战。为了能准确分类小样本皮肤疾病,提出一种基于空间变换网络(STN)和特征分布校准的小样本分类模型。首先,将迁移学习和元学习相结合,以解决跨域迁移小样本存在的过拟合问题;其次,在预训练分类任务前插入旋转角度预测任务,以便模型更好地适应医学图像数据的高复杂度;再次,在对图像下采样后引入STN,以通过显式地对输入图像进行仿射变换,增强特征的提取和识别能力;最后,通过特征分布校准对新类特征进行约束,并引入最邻近质心算法进行分类决策,在简化算法流程的同时显著提升分类精度。在ISIC2018皮肤病变数据集上的实验结果表明,与当前主流小样本模型Meta-Baseline相比,在2-way和3-way分类任务中,所提模型的平均精度分别提高了11.80和10.82个百分点;与模型MetaMed相比,在2-way 3-shot和3-way 3-shot分类任务中,所提模型的分类精度分别提升了6.65和9.58个百分点。可见,所提模型有效提高了小样本皮肤疾病的分类精度,能够更好地辅助医生提高临床诊断精确度。

关键词: 小样本学习, 图像分类, 皮肤病变, 空间变换网络, 最邻近质心

Abstract:

Deep learning-based image classification methods typically require a lot of labeled data. However, in classification task of skin lesions in the medical field, collecting a lot of image data faces numerous challenges. To classify few-shot skin diseases accurately, a few-shot classification model based on Spatial Transformer Network (STN) and feature distribution calibration was proposed. Firstly, transfer learning and meta-learning were integrated to address the overfitting issue in cross-domain few-shot transfer. Secondly, a rotation angle prediction task was inserted before the pre-training classification task to better adapt the model to the high complexity of medical image data. Thirdly, after downsampling the images, a STN was introduced to perform affine transformations on the input images explicitly, thereby enhancing feature extraction and recognition capabilities. Finally, feature distribution calibration was used to constrain new class features, and the nearest centroid algorithm was introduced for classification decisions, thereby reducing algorithm complexity while improving classification accuracy significantly. Experimental results on ISIC2018 skin lesion dataset show that compared to the current mainstream few-shot model Meta-Baseline, the proposed model has the accuracy improvements of 11.80 and 10.82 percentage points in 2-way and 3-way classification tasks, respectively; compared to the model MetaMed, the proposed model has the average accuracy improvements of 6.65 and 9.58 percentage points in 2-way 3-shot and 3-way 3-shot classification tasks, respectively. It can be seen that the proposed model improves the classification accuracy of few-shot skin diseases effectively, and can assist doctors better in enhancing clinical diagnosis accuracy.

Key words: few-shot learning, image classification, skin lesion, Spatial Transformer Network (STN), nearest centroid

中图分类号: