OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Image & Video AI

Fbcnn

FBCNN is an open-source Python and PyTorch research implementation of a blind JPEG artifact removal model that restores grayscale and color images, including double-compressed and real-world JPEGs.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Fbcnn — OpenViST project profile in Image & Video AI.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use FBCNN when working with JPEG restoration where the compression quality factor may be unknown, or when exploring the model's documented training and test scripts.

Who it is for

  • Researchers and developers working on image restoration or JPEG artifact removal.
  • Users who need a documented model workflow for grayscale, color, double-JPEG, or real-world JPEG images.

Documented uses

  • Reduce visible artifacts in JPEG-compressed images.
  • Test restoration workflows on grayscale, color, double-compressed, or real-world JPEGs.
  • Train the model using the repository's documented training script and JSON configuration files.
Before you choose

Requirements & limits

The public repository is open source under Apache-2.0. Its README links to a Gradio web demo hosted on Hugging Face Spaces as a separate way to try JPEG images.

  • The README specifies Python 3.7+ and PyTorch 1.7.
  • Separate test scripts are listed for grayscale, double-JPEG grayscale, color, and real-world color images.
  • Training uses main_train_fbcnn.py and JSON configuration files in ./options/.
  • The model's predicted quality factor is used to control the trade-off between artifact removal and detail preservation.
  • A linked Gradio demo is distinct from the repository's local research-code workflows.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Run a documented FBCNN test script

Select and launch the repository's test script corresponding to a documented JPEG image category.

Before you start

  • Python 3.7 or later
  • PyTorch 1.7
  • The FBCNN repository

Steps

  1. Choose the test script for grayscale, double-JPEG grayscale, color, or real-world color JPEG images.

  2. Run the corresponding documented command: the command below, the command below, the command below, or the command below.

    Documented command
    python main_test_fbcnn_gray.py
    Documented command
    python main_test_fbcnn_gray_doublejpeg.py
    Documented command
    python main_test_fbcnn_color.py
    Documented command
    python main_test_fbcnn_color_real.py

Expected outcome

Starts the documented testing workflow for the selected image category.

Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
16 October 2026
Original website title
jiaxi-jiang/FBCNN · Official homepage title · retained in its source language
How we review sourcesSuggest a correction