fbcnn#

Use Case and High-Level Description#

The fbcnn model is a flexible blind convolutional neural network to remove JPEG artifacts. Model based on “Towards Flexible Blind JPEG Artifacts Removal” paper. It was implemented in PyTorch* framework. Model works with color jpeg images. For details about this model and other jpeg artifacts removal models (for grayscale images and double jpeg restoration), check out the “Towards Flexible Blind JPEG Artifacts Removal (FBCNN, ICCV 2021)”.

Specification#

Metric

Value

Type

Image Processing

GFLOPs

1420.78235

MParams

71.922

Source framework

PyTorch*

Accuracy#

Model was tested on LIVE_1 dataset.

Metric

Original model

Converted model

PSNR

34.34Db

34.34Db

SSIM

0.99

0.99

Input#

Original model#

Image, name - image_lq, shape - 1, 3, 512, 512, format is B, C, H, W, where:

  • B - batch size

  • C - channel

  • H - height

  • W - width

Channel order is RGB. Scale value - 255.

Converted model#

Image, name - image_lq, shape - 1, 3, 512, 512, format is B, C, H, W, where:

  • B - batch size

  • C - channel

  • H - height

  • W - width

Channel order is BGR

Output#

Original Model#

Restored image, name - image_result, shape - 1, 3, 512, 512, output data format is B, C, H, W, where:

  • B - batch size

  • C - channel

  • H - height

  • W - width

Channel order is RGB.

Converted Model#

Restored image, name - image_result, shape - 1, 3, 512, 512, output data format is B, C, H, W, where:

  • B - batch size

  • C - channel

  • H - height

  • W - width

Channel order is BGR.

Download a Model and Convert it into OpenVINO™ IR Format#

You can download models and if necessary convert them into OpenVINO™ IR format using the Model Downloader and other automation tools as shown in the examples below.

An example of using the Model Downloader:

omz_downloader --name <model_name>

An example of using the Model Converter:

omz_converter --name <model_name>

Demo usage#

The model can be used in the following demos provided by the Open Model Zoo to show its capabilities: