Show HN: Image-to-Image Translation Model
5 ykjs 1 5/28/2025, 11:47:40 PM jigsawstack.com ↗
We launched a v1 of a image to image translation API which translates the text on an images by replacing the existing text.
For v1, it's pretty much a model pipeline: OCR current text -> generate mask -> erase text -> translate text -> use embedding comparison to find similar font -> map text back on image
v1 was more like a prototype which already beats many of the similar services provided by Google, Azure, etc
We're working on v2 where we're training a diffusion model to translate the text on the image. We've got the pipeline working for English and Chinese, and now we're building datasets for other languages.
Comments (1)
lunaps · 22h ago
thats a great very niched idea. the results seem to go in a promising direction. i think more languages will make the product much more interesting.