Pretraining is All You Need for Image-to-Image Translation
This work addresses image-to-image translation for computer vision applications, offering a generic solution that improves quality but is incremental as it builds on existing pretrained models.
The paper tackles the problem of high-quality image-to-image translation with limited paired data by proposing a framework that adapts pretrained diffusion models, achieving unprecedented realism and faithfulness in synthesized images across benchmarks like ADE20K, COCO-Stuff, and DIODE.
We propose to use pretraining to boost general image-to-image translation. Prior image-to-image translation methods usually need dedicated architectural design and train individual translation models from scratch, struggling for high-quality generation of complex scenes, especially when paired training data are not abundant. In this paper, we regard each image-to-image translation problem as a downstream task and introduce a simple and generic framework that adapts a pretrained diffusion model to accommodate various kinds of image-to-image translation. We also propose adversarial training to enhance the texture synthesis in the diffusion model training, in conjunction with normalized guidance sampling to improve the generation quality. We present extensive empirical comparison across various tasks on challenging benchmarks such as ADE20K, COCO-Stuff, and DIODE, showing the proposed pretraining-based image-to-image translation (PITI) is capable of synthesizing images of unprecedented realism and faithfulness.