Turning Diffusion-Based Image Colorization Into Efficient Color Compression
Pascal Peter, Lilli Kaufhold, Joachim Weickert
The work of Levin et al. (2004) popularized stroke-based methods that add color to gray value images according to a small amount of user-specified color samples. Even though such reconstructions from sparse data suggest a possible use in compression, only few attempts were made so far in this direction. Diffusion-based compression methods pursue a similar idea: they store only few image pixels and inpaint the missing regions. Despite this close relation and a lack of diffusion-based color codecs, colorization ideas were so far only integrated into transform-based approaches such as JPEG. We address this missing link with two contributions. First, we show the relation between the discrete colorization of Levin et al. and continuous diffusion-based inpainting in the YCbCr color space. It decomposes the image into a luma (brightness) channel and two chroma (color) channels. Our luma-guided…