Bit-Depth Enhancement via Convolutional Neural Network
- Install TensorFlow(GPU);
- Run 4-16/test_416.py to recover 16-bit images from 4-bit versions.
Run 8-16/test_816.py to recover 16-bit images from 8-bit versions.
It will directly compress and reconstruct images from test/. - Results output to results_416/ or results_816/.
- The image size in the code needs to be changed.
If you use this code, please cite the following publication:
Y.Su, W.Sun, J.Liu, G.Zhai, P.Jing, "Photo-realistic Image Bit-depth Enhancement via Residual Transposed Convolutional Neural Network", to appear in NEUROCOMPUTING
This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
For a copy of the GNU General Public License, please see http://www.gnu.org/licenses/.
Here we thanks Christian Ledig et al. who are authors of "Photo-realistic single image super-resolution using a generative adversarial network", published in IEEE Conference of Computer Vision and Pattern Recognition, for referring to their outstanding work.