Text-Aware Image Restoration with Diffusion Models
Jaewon Min, Jin Hyeon Kim, Paul Hyunbin Cho, Jaeeun Lee, Jihye Park, Minkyu Park, Sangpil Kim, Hyunhee Park, Seungryong Kim
2025-06-15
Summary
This paper talks about a new system called Text-Aware Image Restoration (TAIR), which uses advanced technology to improve how broken or unclear images can be fixed, especially when there are words or text in the images.
What's the problem?
The problem is that existing methods have trouble restoring images clearly when those images include text. Sometimes, the text gets messed up or is hard to read after the restoration process.
What's the solution?
The solution was to create a system that combines a special type of image restoration technique called a multi-task diffusion framework with a tool that can detect and focus on text within the images. This way, the system can restore the image while also making sure that the text stays clear and accurate.
Why it matters?
This matters because many images we use daily, like signs, documents, or photos with text, need to be clear for understanding. Improving both the overall image and the text quality in restored images makes the technology more useful and reliable.
Abstract
The proposed Text-Aware Image Restoration (TAIR) system integrates a multi-task diffusion framework with a text-spotting module to enhance both image recovery and textual fidelity, outperforming existing diffusion-based methods.