- Name output nodes of models
- Add support for multiple outputs in models and in training
- Update loss output format for cli and gui
- Sort graphs tabs on training
- Fix analysis graph for new loss names
- Multi output support in convert
Tool: Add tool to restore models from backup
Snapshot: Create snapshot based on total iterations rather than session iterations
Models: Move backup/snapshot functions to lib/model
Training: Output average loss since last save at each save iteration
GUI: display_page.py: minor logging update
- Add error handling to LongRunningTask
- Analysis: Fix timestats bug for missing session data
- Training: Update training graph inside LongRunningTask
- Training: Smoothing Amount Slider geometry fix
- Training: Auto update Analysis tab on training start and save interval
- Scripts: Keep control of the UI when terminating by terminating using LongRunningTask
- Analysis: Handle missing session Ids
- Analysis: Handle invalid data selections
- Analysis: Don't update graph on check box selection
- Analysis: Add cursor events to analysis tab
- Analysis: Change default graph view
- Training: Prevent GUI from attempting to close graphs which do not exist
- Extract/Convert: Make file extensions case-insensitive for preview
- Progress bar: Make compatible for downloads
- Progress bar: Add TQDM Description to progress bar stats text
* PlaidML support in setup.py
* Add AMD flag for plaidml support
* PlaidML - Use .plaidML file if present, otherwise default to largest available
* Update training info message for GUI
* Add AMD option to Windows Installer
* Add quiet keras.backend importer and minor sysinfo amend.
* plug PlaidMLStats into GPUStats
* Update Docs
Remove FFMPEG from Conda Requirements
Fix Tensorboard logging for Tensorflow 1.13.1
Bump default Conda Tensorflow to 1.13.1
Update max python version to 3.7
Bugfix: Fully disable keypress monitor for GUI
Bugfix: Preview - Handle missing alignments file
Config changes:
- Separate plugin defaults into their own files
- Move mask_type to global training config
- Add ability to pass in custom config files
* Color Augmentation: Implement testing code
* Tweak Contrast and Lighting augmentation amounts
* Prevent color augmentation from entering timelapse
* Remove all augmentations from preview images
Convert - Force prediction batchsize to 1 when converting on CPU
Training: Move snapshot saving to outside of the save threads to avoid race condition
GUI Summary stats. Show iteration count for current training session (fixed to last save amount).
bug in extract.py: because of missing parentheses in a modulo operation, a condition on save-interval was never met. alignements.json
file was never generated when starting extract script with -si option.
sysinfo,py: CUDNN Check fix for unsupported OSes
cli.py: More useful information
scripts/convert.py: Set processes to 1 if no images are passed in
cv2_dnn aligner: Fix bounding box for negative values
scripts/convert.py: Fix for incorrect frame ranges passed in
* Sort to vgg_face. Remove Face Recognition requirements
* Remove dlib components from setup.py
* Remove dlib from Windows Installer script
* Remove dlib and cmake from requirements.txt
* Update INSTALL.md to remove all dlib requirements
* Remove dlib and cmake from dockerfiles
* Windows installer. Output cli to details Window. Setup.py bugfix
* Skip FaceFilter if invalid paths passed in
* Implement extraction pipeline
* Face filter to vgg_face. Resume partial model downloads
* On-the-fly conversion to extraction pipeline
* Move git model ids from get_model to model definition
* Separate predict and implement pool
* Add check and raise error to multithreading
Box functions to config. Add crop box option.
* All masks to mask module. Refactor convert masks
Predicted mask passed from model. Cli update
* Intesect box with mask and fixes
* Use raw NN output for convert
Use raw mask for face adjustments. Split adjustments to pre and post warp
* Separate out adjustments. Add unmask sharpen
* Set sensible defaults. Pre PR Testing
* Fix queue sizes. Move masked.py to lib
* Fix Skip Frames. Fix GUI Config popup
* Sensible queue limits. Add a less resource intensive single processing mode
* Fix predicted mask. Amend smooth box defaults
* Deterministic ordering for video output
* Video to Video convert implemented
* Fixups
- Remove defaults from folders across all stages
- Move match-hist and aca into color adjustments selectable
- Handle crashes properly for pooled processes
- Fix output directory does not exist error when creating a new output folder
- Force output to frames if input is not a video
* Add Color Transfer adjustment method
Wrap info text in GUI plugin configure popup
* Refactor image adjustments. Easier to create plugins
Start implementing config options for video encoding
* Add video encoding config options
Allow video encoding for frames input (must pass in a reference video)
Move video and image output writers to plugins
* Image writers config options
Move scaling to cli
Move draw_transparent to images config
Add config options for cv2 writer
Add Pillow image writer
* Add gif filetype to Pillow. Fix draw transparent for Pillow
* Add Animated GIF writer
standardize opencv/pillow defaults
* [speedup] Pre-encode supported writers in the convert pool (opencv, pillow)
Move scaling to convert pool
Remove dfaker mask
* Fix default writer
* Bugfixes
* Better custom argparse formatting
torzdf