I2I-Galip: Unsupervised Medical Image Translation Using Generative Adversarial CLIP

TL;DR

I2I-Galip introduces an unsupervised, multi-domain medical image translation framework leveraging a pre-trained CLIP model, enabling efficient translation with a single lightweight generator, outperforming traditional methods like CycleGAN.

Contribution

The paper presents a novel framework that uses CLIP to reduce the need for multiple generator-discriminator pairs in multi-domain image translation.

Findings

Outperforms existing methods on MRI and CT datasets

Uses a single generator with ~13M parameters for multiple domains

Achieves better translation quality and efficiency

Abstract

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).