A Fast and Accurate Super-Resolution Convolutional Neural Network (FSRCNN) build for artwork, anime, and illustration.
Enlargement and Improvement of Anime Images using Modified Fast and Accurate Super Resolution Convolutional Neural Network
This project and paper were officially canceled by the creator on April 29, 2021, due to the physical permanent resident move of the project creators and the COVID-19 pandemic.
https://github.com/TanakitInt/SRCNN-anime
If you need to use any of my Drafted Paper or material, please kindly reference us.
In the future, both SRCNN-anime (1) and FSRCNN-anime (2) trained model will be used for Web Application instead, scheduled on June 2021. But due to the graduation of the project creators, the scheduled Web Application would be postponed until future notice.
(1) Modified SRCNN model
(2) SRCNN model, FSRCNN model
Trained Model (1):
https://drive.google.com/drive/folders/1m27aL3ZZsIlG3ahuZnMsubPYNa2QdfUH?usp=sharing
Trained Model (2):
https://drive.google.com/drive/folders/176VqeUGuNToDu8rnCZ9tpvW3Rms3prMX?usp=sharing
h5 file:
https://drive.google.com/drive/folders/1S3Htyhmv0gPEUjfKWrzg4CCW67SJm388?usp=sharing
Presentation:
https://drive.google.com/file/d/1mw3rW_Xr36Pqw8Ws2GyqPFHnpuELES7d/view?usp=sharing
Drafted paper:
https://drive.google.com/file/d/1gX8vYJ_MGB8LC0ytG_1V-Wux5y_HrmQV/view?usp=sharing
Tanakit Intaniyom
May 5, 2021