2018/09/17 by Jae-Hyeok Lee, Seong Tae Kim, Lee, Jae-Hyeok +5
Biochemistry, Genetics and Molecular Biology · Computer Science · #AI in cancer detection #Cell Image Analysis Techniques #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Generative Adversarial Networks and Image Synthesis
paper · pdf · doi:10.48550/arxiv.1809.06147
openalex publication_date 2018/09/17 · openalex created_date 2022/08/03 · openalex updated_date 2026/07/28
This paper deals with a method for generating realistic labeled masses.\nRecently, there have been many attempts to apply deep learning to various\nbio-image computing fields including computer-aided detection and diagnosis. In\norder to learn deep network model to be well-behaved in bio-image computing\nfields, a lot of labeled data is required. However, in many bioimaging fields,\nthe large-size of labeled dataset is scarcely available. Although a few\nresearches have been dedicated to solving this problem through generative\nmodel, there are some problems as follows: 1) The generated bio-image does not\nseem realistic; 2) the variation of generated bio-image is limited; and 3)\nadditional label annotation task is needed. In this study, we propose a\nrealistic labeled bio-image generation method through visual feature processing\nin latent space. Experimental results have shown that mass images generated by\nthe proposed method were realistic and had wide expression range of targeted\nmass characteristics.\n