Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{199068,
        author = {Pritish Priyadarshi Patra and Prakhar Sahu and Kshitij Prasad and Punith Kumar H and Vibha T G},
        title = {Data Hiding Using Style Transfer and Diffusion},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {12641-12645},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=199068},
        abstract = {Traditional image hiding often leaves signs. Changes in pixels or frequency can create distortions that can be spotted by methods. This work checks if information can be hidden well so that advanced tools can't find it. To do this we suggest a two-layer approach. It combines neural style transfer and diffusion-based inpainting to hide information. Secret data, like text or images is first turned into a grid using a code. This grid then goes through a VGG-19 style transfer model. The cover image adds its patterns to the grid. The information gets embedded at a feature level not by changing pixels A second stage uses Stable Diffusion inpainting to remove style transfer artifacts. It rebuilds changed areas in latent space with a mask and prompt. The final images look natural and consistent. Tests show the results of our method. The SSIM is up to 0.78. The PSNR is 30 dB. The LPIPS is below 0.18. Our method is strong, against attacks. It hides information well.},
        keywords = {Steganography, Neural Style Transfer, Diffusion Inpainting, VGG-19},
        month = {April},
        }