Image Segmentation Mehtod Based on Federal Style Transfer

被引：0
|
作者：
Ma J. [1 ,2 ,3 ,4 ]
Yin X. [1 ,2 ,3 ]
Hu C. [1 ,2 ,5 ]
Ma B. [1 ,2 ,5 ,6 ]
Ban X. [1 ,2 ,5 ,6 ,7 ]
机构：
[1] Beijing Advanced Innovation Center for Materials Genome Engineering, University of Science and Technology Beijing, Beijing
[2] School of Intelligence Science and Technology, University of Science and Technology Beijing, Beijing
[3] Collaborative Innovation Center of Steel Generic Technology, Beijing University of Science and Technology, Beijing
[4] HBIS Group Company Limited, Shijiazhuang
[5] Shunde Innovation School, University of Science and Technology Beijing, Foshan
[6] Key Laboratory of Intelligent Bionic Unmanned Systems, Ministry of Education, University of Science and Technology Beijing, Beijing
[7] Institute for Materials Intelligent Technology, Liaoning Academy of Materials, Shenyang
来源：
Beijing Youdian Daxue Xuebao/Journal of Beijing University of Posts and Telecommunications | 2023年 / 46卷 / 06期
关键词：
deep learning; federated learning; image segmentation; style transfer;
D O I：
10.13190/j.jbupt.2022-257
中图分类号：
学科分类号：
摘要：
In this work, we propose an image segmentation method based on federated style transfer to slove the non-independent and identically distributed problem in federated learning. By sharing style information that is not sensitive to user privacy, this method generates synthetic data for data expansion, and reduces data differences between different users while ensuring that important structural information of data is not disclosed. Experiments results show that this method effectively alleviates the influence of nonindependent and identically distributed problem among nodes on the performance of the federated model in the liver image segmentation task. Therefore, the proposed method can further improve the performance of federal model, which provides the possibility to break the data island and establish a general model in medical field. © 2023 Beijing University of Posts and Telecommunications. All rights reserved.
引用
收藏
页码：27 / 32
页数：5
相关论文
共 21 条
  • [11] HIGAKI T, NAKAMURA Y, TATSUGAMI F, Et al., Improvement of image quality at CT and MRI using deep learning [ J], Japanese Journal of Radiology, 37, 1, pp. 73-80, (2019)
  • [12] SONKA M, HLAVAC V, BOYLE R., Image processing, analysis, and machine vision [ M], pp. 38-45, (2014)
  • [13] MCMAHAN B, MOORE E, RAMAGE D, Et al., Communication-efficient learning of deep networks from decentralized data, 椅Proceedings of Machine Learning Research, pp. 1273-1282, (2017)
  • [14] ACAR D A E, ZHAO Y, NAVARRO R M, Et al., Federated learning based on dynamic regularization
  • [15] LI T, SAHU A K, ZAHEER M, Et al., Federated optimization in heterogeneous networks, Proceedings of Machine Learning and Systems, 2, pp. 429-450, (2020)
  • [16] LIN T, KONG L, STICH S U, Et al., Ensemble distillation for robust model fusion in federated learning, Advances in Neural Information Processing Systems, 33, pp. 2351-2363, (2020)
  • [17] KUNCHEVA L I, WHITAKER C J., Measures of diversity in classifier ensembles and their relationship with the ensemble accuracy, Machine Learning, 51, 2, pp. 181-207, (2003)
  • [18] WANG T C, LIU M Y, ZHU J Y, Et al., High-resolution image synthesis and semantic manipulation with conditional gans, 椅Proceedings of The IEEE Conference on Computer Vision and Pattern Recognition, pp. 8798-8807, (2018)
  • [19] RONNEBERGER O, FISCHER P, BROX T., U-Net: convolutional networks for biomedical image segmentation[ C], 椅 International Conference on Medical Image Computing and Computer-Assisted Intervention, pp. 234-241, (2015)
  • [20] FALK T, MAI D, BENSCH R, Et al., U-Net: deep learning for cell counting, detection, and morphometry, Nature Methods, 16, 1, pp. 67-70, (2019)
123