Nobutaka's Projects Page
Image Super Res utilizes advanced AI technology from Real-ESRGAN to upscale and enhance images. It supports a variety of image types, including photos and anime/illustrations, with scaling options of x2 and x4. Quickly improve the resolution and quality of your photos with the easy-to-use app.
- GIF
- HEIC
- JPEG
- PNG
- WebP
- RealESRGAN x4 Plus: A scale x4 model designed for enhancing general images.
- RealESRGAN x2 Plus: A scale x2 model optimized for upscaling general images with less enlargement.
- RealESRNet x4 Plus: A scale x4 model that uses MSE (Mean Squared Error) loss, ideal for achieving high fidelity in enhanced images.
- realesr general x4v3: A compact scale x4 model, offering a smaller footprint while still providing high-quality image enhancements.
- RealESRGAN x4 Plus Anime: A scale x4 model specifically optimized for anime images, enhancing details while preserving the unique aesthetics of anime art.
For further information, visit the project's GitHub repository: Real-ESRGAN on GitHub.
Citation: Wang, X., Xie, L., Dong, C., & Shan, Y. (2021). Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Data. In Proceedings of the International Conference on Computer Vision Workshops (ICCVW).