Alex J. Champandard
|
881e909dc5
|
Fix default parameters to zoom=2 so it works without arguments. Closes #71.
|
9 years ago |
|
Alex J. Champandard
|
a22d7ced3c
|
Rename folder with training script so they are not used for enhancing by users accidentally.
|
9 years ago |
|
Alex J. Champandard
|
fabe04f1d0
|
Fix bug in handling of --train-jpeg parameter.
|
9 years ago |
|
Alex J. Champandard
|
7c8cc5609f
|
Add training scripts, they take a while to run!
|
9 years ago |
|
Alex J. Champandard
|
9d8ab0ea8e
|
Merge pull request #69 from alexjc/v0.3
Release 0.3
|
9 years ago |
|
Alex J. Champandard
|
80d7b57bec
|
Prepare build scripts for merge and update README.
|
9 years ago |
|
Alex J. Champandard
|
b706599cb1
|
Fix alias for bash and zsh, Closes #54. Add standalone alias-free docker example, Closes #46.
|
9 years ago |
|
Alex J. Champandard
|
c03a0b7fa3
|
Fix for trained models loaded from script directory. Closes #67.
|
9 years ago |
|
Alex J. Champandard
|
a69d2ae060
|
Add histogram color matching, enabled with `--rendering-histogram`.
|
9 years ago |
|
Alex J. Champandard
|
965bd55090
|
Add new model selection (by image type and 'enhance' operation) and parameters and documentation in README.
|
9 years ago |
|
Alex J. Champandard
|
e6120878c6
|
Minor rework of the training parameters.
|
9 years ago |
|
Alex J. Champandard
|
0d4621df0a
|
Simplify architecture outside of convolution blocks for speed, improve training code.
|
10 years ago |
|
Alex J. Champandard
|
c8e805667f
|
Fix for padding in convolution.
|
10 years ago |
|
Alex J. Champandard
|
2a29fc954b
|
Buffer chunks of images proportionally to size.
|
10 years ago |
|
Alex J. Champandard
|
da354f3e55
|
Remove reflection layer, important but not the bottleneck right now and quite slower.
|
10 years ago |
|
Alex J. Champandard
|
9bba9850d5
|
Improve pre-processing steps to include more randomness: scales, blur...
|
10 years ago |
|
Alex J. Champandard
|
149ae4e566
|
Update training parameters.
|
10 years ago |
|
Alex J. Champandard
|
de047d6a74
|
Fix for images that are not multiples of 2 or 4, depending on the rescaling.
|
10 years ago |
|
Alex J. Champandard
|
29bfe58d73
|
Merged documentation changes.
|
10 years ago |
|
Alex J. Champandard
|
9a228c1d78
|
Merge pull request #44 from wieczorek1990/master
Fixed typo.
|
10 years ago |
|
Alex J. Champandard
|
b24fa7490d
|
Merge pull request #47 from alexjc/v0.2
Release 0.2: Tiled Rendering, 1:1 Noise/Blur Fixing
|
10 years ago |
|
Alex J. Champandard
|
03914db364
|
Fix formatting, minor tweaks to Docker build files for release.
|
10 years ago |
|
Alex J. Champandard
|
448e7b93dc
|
Fix progress output when tiled rendering, changed default zoom level.
|
10 years ago |
|
Alex J. Champandard
|
cabaaeeefe
|
Add training scripts for networks currently being trained, for release v0.2.
|
10 years ago |
|
Alex J. Champandard
|
1ad40b6d71
|
Merge branch 'master' into v0.2
|
10 years ago |
|
Alex J. Champandard
|
ac49676415
|
Add tiled rendering with padding, no feather-blending but looks good enough.
|
10 years ago |
Łukasz Wieczorek
|
ed20dcefd5
|
Fixed typo
|
10 years ago |
|
Alex J. Champandard
|
095fe42dc3
|
Add tiled rendering, currently with no padding for each tile.
|
10 years ago |
|
Alex J. Champandard
|
d18c08f1b5
|
Integrated reflection padding instead of zero padding for extra quality during training and inference.
|
10 years ago |
|
Alex J. Champandard
|
90c0b7ea43
|
Fix padding code, more reliable for specific upscale/downscale combinations.
|
10 years ago |
|
Alex J. Champandard
|
3b2a6b9d8d
|
Add extra padding on input to avoid zero-padding. Experiment with training values from ENet (segmentation).
|
10 years ago |
|
Alex J. Champandard
|
7924cc4a85
|
Improve display and filenames for saving output.
|
10 years ago |
|
Alex J. Champandard
|
93e5a41d9a
|
Fix and optimize pre-processing of images.
|
10 years ago |
|
Alex J. Champandard
|
11ba505252
|
Fix for gradient clipping code.
|
10 years ago |
|
Alex J. Champandard
|
cf65207a2e
|
Use full range of tanh output rather than [-0.5, +0.5], avoids clipping.
|
10 years ago |
|
Alex J. Champandard
|
0c31e53731
|
Fix suggested alias for relative paths. Closes #37 #28.
|
10 years ago |
|
Alex J. Champandard
|
34f8e629c2
|
Improve the alias used to invoke docker, so it's more robust to directory locations and input paths.
|
10 years ago |
|
Alex J. Champandard
|
c610623b11
|
Add gradient clipping, helpful for preventing problems with extreme parameters/architectures.
|
10 years ago |
|
Alex J. Champandard
|
02d2fca6c5
|
Corrected value for adversarial loss. Don't refactor math the day after stopping coffee.
|
10 years ago |
|
Alex J. Champandard
|
f2494f8078
|
Add new downscale layers, separate from upscale steps. Renamed --scales to --zoom for inference.
|
10 years ago |
|
Alex J. Champandard
|
0c9937a317
|
Merge pull request #22 from msfeldstein/master
Remove cnmem theano flag.
|
10 years ago |
|
Alex J. Champandard
|
064f9dd589
|
Add three image pre-processing options, improve loading code.
|
10 years ago |
Michael Feldstein
|
fef84c5b44
|
Remove cnmem theano flag since it doesn't work if you're sharing GPU with display.
|
10 years ago |
|
Alex J. Champandard
|
5ef872b876
|
Add warning for files that may be too large for 4x.
|
10 years ago |
|
Alex J. Champandard
|
a5ad2c25e6
|
Merge pull request #18 from alexjc/training
Training Improvements
|
10 years ago |
|
Alex J. Champandard
|
17fcad8d28
|
Refactor of changes related to training.
|
10 years ago |
|
Alex J. Champandard
|
2b67daedb6
|
Merge pull request #12 from dribnet/generic_seeds
Move generation of seeds out of training network.
|
10 years ago |
|
Alex J. Champandard
|
cad5eff572
|
Merge pull request #11 from dribnet/save_every_epoch
Added --save-every-epoch option
|
10 years ago |
|
Alex J. Champandard
|
a9b0cd9887
|
Merge pull request #9 from zuphilip/patch-1
Fix some typos in README.
|
10 years ago |
|
Alex J. Champandard
|
fcc5e87858
|
Merge pull request #4 from dribnet/valid_dir
Add valid dir when necessary.
|
10 years ago |
