I2I-Galip: Unsupervised Medical Image Translation Using Generative Adversarial CLIP
Fuente:
arXiv
Guardado en:
| Autores principales: | , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866929505872904192 |
|---|---|
| author | Korkmaz, Yilmaz Patel, Vishal M. |
| author_facet | Korkmaz, Yilmaz Patel, Vishal M. |
| contents | Unpaired image-to-image translation is a challenging task due to the absence of paired examples, which complicates learning the complex mappings between the distinct distributions of the source and target domains. One of the most commonly used approach for this task is CycleGAN which requires the training of a new pair of generator-discriminator networks for each domain pair. In this paper, we propose a new image-to-image translation framework named Image-to-Image-Generative-Adversarial-CLIP (I2I-Galip) where we utilize a pre-trained multi-model foundation model (i.e., CLIP) to mitigate the need of separate generator-discriminator pairs for each source-target mapping while achieving better and more efficient multi-domain translation. By utilizing the massive knowledge gathered during pre-training a foundation model, our approach makes use of a single lightweight generator network with ~13M parameters for the multi-domain image translation task. Comprehensive experiments on translation performance in public MRI and CT datasets show the superior performance of the proposed framework over the existing approaches. Code will be available (https://github.com/yilmazkorkmaz1/I2I-Galip). |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_12399 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | I2I-Galip: Unsupervised Medical Image Translation Using Generative Adversarial CLIP Korkmaz, Yilmaz Patel, Vishal M. Image and Video Processing Computer Vision and Pattern Recognition Unpaired image-to-image translation is a challenging task due to the absence of paired examples, which complicates learning the complex mappings between the distinct distributions of the source and target domains. One of the most commonly used approach for this task is CycleGAN which requires the training of a new pair of generator-discriminator networks for each domain pair. In this paper, we propose a new image-to-image translation framework named Image-to-Image-Generative-Adversarial-CLIP (I2I-Galip) where we utilize a pre-trained multi-model foundation model (i.e., CLIP) to mitigate the need of separate generator-discriminator pairs for each source-target mapping while achieving better and more efficient multi-domain translation. By utilizing the massive knowledge gathered during pre-training a foundation model, our approach makes use of a single lightweight generator network with ~13M parameters for the multi-domain image translation task. Comprehensive experiments on translation performance in public MRI and CT datasets show the superior performance of the proposed framework over the existing approaches. Code will be available (https://github.com/yilmazkorkmaz1/I2I-Galip). |
| title | I2I-Galip: Unsupervised Medical Image Translation Using Generative Adversarial CLIP |
| topic | Image and Video Processing Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2409.12399 |