A Self-attention Guided Multi-scale Gradient GAN for Diversified X-ray Image Synthesis
نویسندگان
چکیده
Abstract Imbalanced image datasets are commonly available in the domain of biomedical analysis. Biomedical images contain diversified features that significant predicting targeted diseases. Generative Adversarial Networks (GANs) utilized to address data limitation problem via generation synthetic images. Training challenges such as mode collapse, non-convergence, and instability degrade a GAN’s performance synthesizing high-quality In this work, MSG-SAGAN, an attention-guided multi-scale gradient GAN architecture is proposed model relationship between long-range dependencies improves training using flow gradients at multiple resolutions layers generator discriminator models. The intent reduce impact collapse stabilize attention mechanism with learning for X-ray synthesis. Multi-scale Structural Similarity Index Measure (MS-SSIM) Frechet Inception Distance (FID) used identify occurrence evaluate diversity generated. compared (MSG-GAN) assess generated Results indicate MSG-SAGAN outperforms MSG-GAN evidenced by MS-SSIM FID scores.
منابع مشابه
A Multi-Scale Gradient based Method for Image Completion
LU Rui XU De LI Bing Institute of Computer Science and Engineering Beijing Jiaotong University Beijing,100044 CHINA [email protected] [email protected] [email protected] Abstract:Image completion is a challenging problem in computer graphics and computer vision. It refers to the problem of automatically filling holes in images left by the unwanted objects removal in a plausible way. Unlike previou...
متن کاملMR to X-Ray Projection Image Synthesis
Hybrid imaging promises large potential in medical imaging applications. To fully utilize the possibilities of corresponding information from different modalities, the information must be transferable between the domains. In radiation therapy planning, existing methods make use of reconstructed 3D magnetic resonance imaging data to synthesize corresponding Xray attenuation maps. In contrast, fo...
متن کاملfuzzy-based medical x-ray image classification
in this paper a novel fuzzy scheme for medical x-ray image classification is presented. in this method, any image is partitioned in to 25 overlapping subimages and then shape-texture features are extracted from shape and directional information extracted from any subimage. in the classification stage, we apply a fuzzy membership to any subimage with respect to euclidean distance between feature...
متن کاملEvaluation of CUDA for Multi-Energy X-ray Image Reconstruction
High-performance computing capability enabled the development of a multi-energy X-ray imaging system which requires image reconstruction from multi-energy images. Usually, pararell processing of CUDA is faster than CPU. However, the CUDA could not be taking advantage depending on kind of program. In this paper, we investigate the performance of the CUDA for parallel processing in order to evalu...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Communications in computer and information science
سال: 2023
ISSN: ['1865-0937', '1865-0929']
DOI: https://doi.org/10.1007/978-3-031-26438-2_2