jiaxi-jiang/FBCNN
PyTorch implementation of FBCNN, a deep learning model for removing JPEG compression artifacts from images.
Not currently ranked — collecting fresh signals.
star history
FBCNN is a state-of-the-art blind JPEG artifacts removal model published at ICCV 2021. It uses deep convolutional networks to estimate JPEG quality factors and remove compression artifacts from single and double JPEG images, handling both color and grayscale inputs. The model is trained with PyTorch and supports arbitrary quality factor prediction for flexible blind restoration.
Frequently asked
- What is jiaxi-jiang/FBCNN?
- PyTorch implementation of FBCNN, a deep learning model for removing JPEG compression artifacts from images.
- Is FBCNN open source?
- Yes — jiaxi-jiang/FBCNN is open source, released under the Apache-2.0 license.
- What language is FBCNN written in?
- jiaxi-jiang/FBCNN is primarily written in Python.
- How popular is FBCNN?
- jiaxi-jiang/FBCNN has 526 stars on GitHub.
- Where can I find FBCNN?
- jiaxi-jiang/FBCNN is on GitHub at https://github.com/jiaxi-jiang/FBCNN.