* model_refactor (#571)
* original model to new structure
* IAE model to new structure
* OriginalHiRes to new structure
* Fix trainer for different resolutions
* Initial config implementation
* Configparse library added
* improved training data loader
* dfaker model working
* Add logging to training functions
* Non blocking input for cli training
* Add error handling to threads. Add non-mp queues to queue_handler
* Improved Model Building and NNMeta
* refactor lib/models
* training refactor. DFL H128 model Implementation
* Dfaker - use hashes
* Move timelapse. Remove perceptual loss arg
* Update INSTALL.md. Add logger formatting. Update Dfaker training
* DFL h128 partially ported
* Add mask to dfaker (#573)
* Remove old models. Add mask to dfaker
* dfl mask. Make masks selectable in config (#575)
* DFL H128 Mask. Mask type selectable in config.
* remove gan_v2_2
* Creating Input Size config for models
Creating Input Size config for models
Will be used downstream in converters.
Also name change of image_shape to input_shape to clarify ( for future models with potentially different output_shapes)
* Add mask loss options to config
* MTCNN options to config.ini. Remove GAN config. Update USAGE.md
* Add sliders for numerical values in GUI
* Add config plugins menu to gui. Validate config
* Only backup model if loss has dropped. Get training working again
* bugfixes
* Standardise loss printing
* GUI idle cpu fixes. Graph loss fix.
* mutli-gpu logging bugfix
* Merge branch 'staging' into train_refactor
* backup state file
* Crash protection: Only backup if both total losses have dropped
* Port OriginalHiRes_RC4 to train_refactor (OriginalHiRes)
* Load and save model structure with weights
* Slight code update
* Improve config loader. Add subpixel opt to all models. Config to state
* Show samples... wrong input
* Remove AE topology. Add input/output shapes to State
* Port original_villain (birb/VillainGuy) model to faceswap
* Add plugin info to GUI config pages
* Load input shape from state. IAE Config options.
* Fix transform_kwargs.
Coverage to ratio.
Bugfix mask detection
* Suppress keras userwarnings.
Automate zoom.
Coverage_ratio to model def.
* Consolidation of converters & refactor (#574)
* Consolidation of converters & refactor
Initial Upload of alpha
Items
- consolidate convert_mased & convert_adjust into one converter
-add average color adjust to convert_masked
-allow mask transition blur size to be a fixed integer of pixels and a fraction of the facial mask size
-allow erosion/dilation size to be a fixed integer of pixels and a fraction of the facial mask size
-eliminate redundant type conversions to avoid multiple round-off errors
-refactor loops for vectorization/speed
-reorganize for clarity & style changes
TODO
- bug/issues with warping the new face onto a transparent old image...use a cleanup mask for now
- issues with mask border giving black ring at zero erosion .. investigate
- remove GAN ??
- test enlargment factors of umeyama standard face .. match to coverage factor
- make enlargment factor a model parameter
- remove convert_adjusted and referencing code when finished
* Update Convert_Masked.py
default blur size of 2 to match original...
description of enlargement tests
breakout matrxi scaling into def
* Enlargment scale as a cli parameter
* Update cli.py
* dynamic interpolation algorithm
Compute x & y scale factors from the affine matrix on the fly by QR decomp.
Choose interpolation alogrithm for the affine warp based on an upsample or downsample for each image
* input size
input size from config
* fix issues with <1.0 erosion
* Update convert.py
* Update Convert_Adjust.py
more work on the way to merginf
* Clean up help note on sharpen
* cleanup seamless
* Delete Convert_Adjust.py
* Update umeyama.py
* Update training_data.py
* swapping
* segmentation stub
* changes to convert.str
* Update masked.py
* Backwards compatibility fix for models
Get converter running
* Convert:
Move masks to class.
bugfix blur_size
some linting
* mask fix
* convert fixes
- missing facehull_rect re-added
- coverage to %
- corrected coverage logic
- cleanup of gui option ordering
* Update cli.py
* default for blur
* Update masked.py
* added preliminary low_mem version of OriginalHighRes model plugin
* Code cleanup, minor fixes
* Update masked.py
* Update masked.py
* Add dfl mask to convert
* histogram fix & seamless location
* update
* revert
* bugfix: Load actual configuration in gui
* Standardize nn_blocks
* Update cli.py
* Minor code amends
* Fix Original HiRes model
* Add masks to preview output for mask trainers
refactor trainer.__base.py
* Masked trainers converter support
* convert bugfix
* Bugfix: Converter for masked (dfl/dfaker) trainers
* Additional Losses (#592)
* initial upload
* Delete blur.py
* default initializer = He instead of Glorot (#588)
* Allow kernel_initializer to be overridable
* Add ICNR Initializer option for upscale on all models.
* Hopefully fixes RSoDs with original-highres model plugin
* remove debug line
* Original-HighRes model plugin Red Screen of Death fix, take #2
* Move global options to _base. Rename Villain model
* clipnorm and res block biases
* scale the end of res block
* res block
* dfaker pre-activation res
* OHRES pre-activation
* villain pre-activation
* tabs/space in nn_blocks
* fix for histogram with mask all set to zero
* fix to prevent two networks with same name
* GUI: Wider tooltips. Improve TQDM capture
* Fix regex bug
* Convert padding=48 to ratio of image size
* Add size option to alignments tool extract
* Pass through training image size to convert from model
* Convert: Pull training coverage from model
* convert: coverage, blur and erode to percent
* simplify matrix scaling
* ordering of sliders in train
* Add matrix scaling to utils. Use interpolation in lib.aligner transform
* masked.py Import get_matrix_scaling from utils
* fix circular import
* Update masked.py
* quick fix for matrix scaling
* testing thus for now
* tqdm regex capture bugfix
* Minor ammends
* blur size cleanup
* Remove coverage option from convert (Now cascades from model)
* Implement convert for all model types
* Add mask option and coverage option to all existing models
* bugfix for model loading on convert
* debug print removal
* Bugfix for masks in dfl_h128 and iae
* Update preview display. Add preview scaling to cli
* mask notes
* Delete training_data_v2.py
errant file
* training data variables
* Fix timelapse function
* Add new config items to state file for legacy purposes
* Slight GUI tweak
* Raise exception if problem with loaded model
* Add Tensorboard support (Logs stored in model directory)
* ICNR fix
* loss bugfix
* convert bugfix
* Move ini files to config folder. Make TensorBoard optional
* Fix training data for unbalanced inputs/outputs
* Fix config "none" test
* Keep helptext in .ini files when saving config from GUI
* Remove frame_dims from alignments
* Add no-flip and warp-to-landmarks cli options
* Revert OHR to RC4_fix version
* Fix lowmem mode on OHR model
* padding to variable
* Save models in parallel threads
* Speed-up of res_block stability
* Automated Reflection Padding
* Reflect Padding as a training option
Includes auto-calculation of proper padding shapes, input_shapes, output_shapes
Flag included in config now
* rest of reflect padding
* Move TB logging to cli. Session info to state file
* Add session iterations to state file
* Add recent files to menu. GUI code tidy up
* [GUI] Fix recent file list update issue
* Add correct loss names to TensorBoard logs
* Update live graph to use TensorBoard and remove animation
* Fix analysis tab. GUI optimizations
* Analysis Graph popup to Tensorboard Logs
* [GUI] Bug fix for graphing for models with hypens in name
* [GUI] Correctly split loss to tabs during training
* [GUI] Add loss type selection to analysis graph
* Fix store command name in recent files. Switch to correct tab on open
* [GUI] Disable training graph when 'no-logs' is selected
* Fix graphing race condition
* rename original_hires model to unbalanced
* Add ability to extract from and convert from a video file
* Update cli helptext. Add filebrowser button for GUI input
* Add video support to Alignments Tool
* add new cli options
* add timelapse script
* add batch size on image generation
* fix circular dependency
* add attribute check
* refactor to 3 parameters
* fix some typos
* do not shuffle the set
* default to models folder when no timelapse output supplied
* almost forgot this
* Add preview functionality to effmpeg. (#435)
* Add preview functionality to effmpeg.
effmpeg tool:
Preview for actions that have a video output now available.
Preview does not work when muxing audio.
* Model json unicode fix1 (#443)
* fixed Windows 10 path error while loading weights
* - fixed TypeError: the JSON object must be str, not 'bytes' with OriginalHighRes Model
* MTCNN Extractor and Extraction refactor (#453)
* implement mtcnn extractor
* mtcnn refactor and vram management changes
* cli arguments update for mtcnn/dlib split
* Add mtcnn models to gitignore
* Change multiprocessing on extract
* GUI changes to handle nargs defaults
* Early exit bugfix (#455)
* Fix extract early termination bug
* Fix extract early exit bug
* Multi face detection bugfix (#456)
* Multi face extraction fix
* Original high res cleanup 1 (#457)
* slight model re-factoring
- removed excess threading code
- added random kernel initialization to dense layer
* Slight OriginalHighRes re-factoring an code cleanup
* GUI version 3 (#411)
GUI version 3.0a
* Required for Shaonlu mode (#416)
Added two modes - Original and Shaonlu.
The later requires this file to function.
* model update (#417)
New, functional Original 128 model
* OriginalHighRes 128 model update (#418)
Required for OriginalHighRes Model to function
* Add OriginalHighRes 128 update to gui branch (#421)
* Required for Shaonlu mode (#416)
Added two modes - Original and Shaonlu.
The later requires this file to function.
* model update (#417)
New, functional Original 128 model
* OriginalHighRes 128 model update (#418)
Required for OriginalHighRes Model to function
* Dev gui (#420)
* reduce singletons
* Fix tooltips and screen boundaries on popup
* Remove dpi fix. Fix context filebrowsers
* fix tools.py execution and context filebrowser bugs
* Bugfixes (#422)
* Bump matplotlib requirement. Fix polyfit. Fix TQDM on sort
* Fixed memory usage at 6GB cards. (#423)
- Switched default encoder to ORIGINAL
- Fixed memory consumption. Tested with geforce gtx 9800 ti with 6Gb; batch_size 8 no OOM or memory warnings now.
* Staging (#426)
* altered trainer (#425)
altered trainer to accommodate with model change
* Update Model.py (#424)
- Added saving state (currently only saved epoch number, to be extended in future)
- Changed saving to ThreadPoolExecutor
* Add DPI Scaling (#428)
* Add dpi scaling
* Hotfix for effmpeg. (#429)
effmpeg fixed so it works both in cli and gui.
Initial work done to add previewing feature to effmpeg (currently does nothing).
Some small spacing changes in other files to improve PEP8 conformity.
* PEP8 Linting (#430)
* pep8 linting
* Requirements version bump (#432)
* altered trainer (#425)
altered trainer to accommodate with model change
* Update Model.py (#424)
- Added saving state (currently only saved epoch number, to be extended in future)
- Changed saving to ThreadPoolExecutor
* Requirements version bump (#431)
This bumps the versions of:
scandir
h5py
Keras
opencv-python
to their latest vesions.
Virtual Environment will need to be setup again to make use of these.
* High DPI Fixes (#433)
* dpi scaling
* DPI Fixes
* Fix and improve context manager. (#434)
effmpeg tool:
Context manager for GUI fixed.
Context manager in general:
Functionality extended to allow configuring the context with both:
command -> action
command -> variable (cli argument) -> action
* Change epoch option to iterations
* Change epochs to iterations
All:
Set correct python3 shebang.
lib/cli.py:
Fix some help documentation formatting and typos.
Set 'json' as the default value for '--serializer' argument.
lib/Serializer.py:
Refactor to properly handle PyYAML not being available.
Add docstring at top of the file.
Improve PEP8 conformity.
scripts/fsmedia.py:
Modify lib/Serializer.py method call to match new name.
Modify lib/Serializer.py method call to match not needing a default.
tools/sort.py:
Add new group by 'face-yaw' method.
Add docstring at top of the file.
Re-arrange argument order to make more sense.
Fix typos and line length issues in help documentation.
Change to use lib/Serializer.py to set the serializer and to write the log file.
Allow sort logging to use PyYAML.
* Pre push commit.
Add filetypes support to gui through new classes in lib/cli.py
Add various new functions to tools/effmpeg.py
* Finish developing basic effmpeg functionality.
Ready for public alpha test.
* Add ffmpy to requirements.
Fix gen-vid to allow specifying a new file in GUI.
Fix extract throwing an error when supplied with a valid directory.
Add two new gui user pop interactions: save (allows you to create new
files/directories) and nothing (disables the prompt button when it's not
needed).
Improve logic and argument processing in effmpeg.
* Fix post merge bugs.
Reformat tools.py to match the new style of faceswap.py
Fix some whitespace issues.
* Fix matplotlib.use() being called after pyplot was imported.
* Fix various effmpeg bugs and add ability do terminate nested subprocess
to GUI.
effmpeg changes:
Fix get-fps not printing to terminal.
Fix mux-audio not working.
Add verbosity option. If verbose is not specified than ffmpeg output is
reduced with the -hide_banner flag.
scripts/gui.py changes:
Add ability to terminate nested subprocesses, i.e. the following type of
process tree should now be terminated safely:
gui -> command -> command-subprocess
-> command-subprocess -> command-sub-subprocess
* Add functionality to tools/effmpeg.py, fix some docstring and print statement issues in some files.
tools/effmpeg.py:
Transpose choices now display detailed name in GUI, while in cli they can
still be entered as a number or the full command name.
Add quiet option to effmpeg that only shows critical ffmpeg errors.
Improve user input handling.
lib/cli.py; scripts/convert.py; scripts/extract.py; scripts/train.py:
Fix some line length issues and typos in docstrings, help text and print statements.
Fix some whitespace issues.
lib/cli.py:
Add filetypes to '--alignments' argument.
Change argument action to DirFullPaths where appropriate.
* Bug fixes and improvements to tools/effmpeg.py
Fix bug where duration would not be used even when end time was not set.
Add option to specify output filetype for extraction.
Enchance gen-vid to be able to generate a video from images that were zero padded to any arbitrary number, and not just 5.
Enchance gen-vid to be able to use any of the image formats that a video can be extracted into.
Improve gen-vid output video quality.
Minor code quality improvements and ffmpeg argument formatting improvements.
* Remove dependency on psutil in scripts/gui.py and various small improvements.
lib/utils.py:
Add _image_extensions and _video_extensions as global variables to make them easily portable across all of faceswap.
Fix lack of new lines between function and class declarions to conform to PEP8.
Fix some typos and line length issues in doctsrings and comments.
scripts/convert.py:
Make tqdm print to stdout.
scripts/extract.py:
Make tqdm print to stdout.
Apply workaround for occasional TqdmSynchronisationWarning being thrown.
Fix some typos and line length issues in doctsrings and comments.
scripts/fsmedia.py:
Did TODO in scripts/fsmedia.py in Faces.load_extractor(): TODO Pass extractor_name as argument
Fix lack of new lines between function and class declarions to conform to PEP8.
Fix some typos and line length issues in doctsrings and comments.
Change 2 print statements to use format() for string formatting instead of the old '%'.
scripts/gui.py:
Refactor subprocess generation and termination to remove dependency on psutil.
Fix some typos and line length issues in comments.
tools/effmpeg.py
Refactor DataItem class to use new lib/utils.py global media file extensions.
Improve ffmpeg subprocess termination handling.
* Refactor for PEP 8 and split process function
* Remove backwards compatibility for skip frames
* Split optional functions into own class. Make functions more modular
* train.py - Fix write image bug. Make more modular
* convert.py: Extract alignments from frames if they don't exist
* BugFix: SkipExisting broken since face name refactor
* train.py - Semi-fix for hang on reaching target iteration. Now quits on preview mode
Make tensorflow / system warning less verbose
* Final bugfixes
* Add 'all' back in for selectable detectors
* Final minor tweaks
* Refactor for PEP 8 and split process function
* Remove backwards compatibility for skip frames
* Split optional functions into own class. Make functions more modular
* Conform scripts folder to PEP 8
* train.py - Fix write image bug. Make more modular
* extract.py - Make more modular, Put optional actions into own class
* cli.py - start PEP 8
* cli,py - Pep 8. Refactor and make modular. Bugfixes
* 1st round refactor. Completely untested and probably broken.
* convert.py: Extract alignments from frames if they don't exist
* BugFix: SkipExisting broken since face name refactor
* Extract.py tested
* Minor formatting
* convert.py + train.py amended not tested
* train.py - Semi-fix for hang on reaching target iteration. Now quits on preview mode
Make tensorflow / system warning less verbose
* 2nd pass refactor. Semi tested
bugfixes
* Remove obsolete code. imread/write to Utils
* rename inout.py to fsmedia.py
* Final bugfixes
* Move imports to top and add conditions to load modules requiring GPU
* Fix type
* Fix typo
* Import relocation and dlib import on demand
* refactor to remove extra code lines
* remove unnecessary sys import
* Add support for user-specified rotation angle in extract
* Added rotation-angle-list option to enumerate a list of angles to rotate through
* Adjust rotation matrix translation coords to avoid cropping
* Merged rotation-angle and rotation-angle-list options into rotate_images option
* Backwards compatibility
* Updated check whether to run image rotator
* Switched rotation convention to use positive angle = clockwise rotation, for backwards compatibility
* Improving performance of extraction. Two main changes to improve the most recent modifications to extract: 1st FaceLandmarkExtractor would try to use cnn first, then try hog. The problem was that this reduced the speed by 4 for images where cnn didn't find anything, and most of the times hog wouldn't find anything either or it would be a bad extract. For me it wasn't worth it. With this you can specify on input -D if you want to use hog, cnn, or all. 'all' will try cnn, then hog like FaceLandmarkExtractor was doing. cnn or hog will just use 1 detection method. 2nd change is a rehaul of the verbose parameter. Now warnings when a face is not detected will just be shown if indicated by -v or --verbose. This restores the verbose function to what it once was. With this change I was able to process 1,000 per each 4 minutes regardless if faces were detected or not. Performance improvement just applies to not detected images but I normally will have lots of images without clear faces in my set, so I figured it would impact others. Also the introduction of 'all' would allow trying other models together more easily in the future.
* Update faces_detect.py
* Update extract.py
* Update FaceLandmarksExtractor.py
* spacing fix
* Image rotator for extract and convert ready for testing
* Revert "Image rotator for extract and convert ready for testing"
This reverts commit bbeb19ef26.
Error in extract code
* add image rotation support to detect more faces
* Update convert.py
Amended to do a single check for for rotation rather than checking twice. Performance gain is likely to be marginal to non-existent, but can't hurt.
* Update convert.py
remove type
* cli.py: Only output message on verbose. Convert.py: Only check for rotation amount once
* Changed command line flag to take arguments to ease future development
* Realigning for upstream/Master
* Minor fix
* Add negative filters for face detection
When detecting faces that are very similar, the face recognition can
produce positive results for similar looking people. This commit allows
the user to add multiple positive and negative reference images. The
facedetection then calculates the distance to each reference image
and tries to guess which is more likely using the k-nearest method.
* Do not calculate knn if no negative images are given
* Clean up outputting
* Clearer requirements for each platform
* Refactoring of old plugins (Model_Original + Extract_Align) + Cleanups
* Adding GAN128
* Update GAN to v2
* Create instance_normalization.py
* Fix decoder output
* Revert "Fix decoder output"
This reverts commit 3a8ecb8957.
* Fix convert
* Enable all options except perceptual_loss by default
* Disable instance norm
* Update Model.py
* Update Trainer.py
* Match GAN128 to shaoanlu's latest v2
* Add first_order to GAN128
* Disable `use_perceptual_loss`
* Fix call to `self.first_order`
* Switch to average loss in output
* Constrain average to last 100 iterations
* Fix math, constrain average to intervals of 100
* Fix math averaging again
* Remove math and simplify this damn averagin
* Add gan128 conversion
* Update convert.py
* Use non-warped images in masked preview
* Add K.set_learning_phase(1) to gan64
* Add K.set_learning_phase(1) to gan128
* Add missing keras import
* Use non-warped images in masked preview for gan128
* Exclude deleted faces from conversion
* --input-aligned-dir defaults to "{input_dir}/aligned"
* Simplify map operation
* port 'face_alignment' from PyTorch to Keras. It works x2 faster, but initialization takes 20secs.
2DFAN-4.h5 and mmod_human_face_detector.dat included in lib\FaceLandmarksExtractor
fixed dlib vs tensorflow conflict: dlib must do op first, then load keras model, otherwise CUDA OOM error
if face location not found by CNN, its try to find by HOG.
removed this:
- if face.landmarks == None:
- print("Warning! landmarks not found. Switching to crop!")
- return cv2.resize(face.image, (size, size))
because DetectedFace always has landmarks
* Enabled masked converter for GAN models
* Histogram matching, cli option for perceptual loss
* Fix init() positional args error
* Add backwards compatibility for aligned filenames
* Fix masked converter
* Remove GAN converters
* Pytorch and face-alignment
* Skip processed frames when extracting faces.
* Reset to master version
* Reset to master
* Added --skip-existing argument to Extract script. Default is to NOT skip already processed frames.
Added logic to write_alignments to append new alignments (and preserve existing ones)
to existing alignments file when the skip-existing option is used.
* Fixed exception for --skip-existing when using the convert script
* Sync with upstream
* Fixed error when using Convert script.
* Bug fix
* Merges alignments only if --skip-existing is used.
* Creates output dir when not found, even when using --skip-existing.
* Preparing GAN plugin
* Adding multithreading for extract
* Adding support for mmod human face detector
* Adding face filter argument
* Added process number argument to multiprocessing extractor.
Fixed progressbar for multiprocessing.
* Added tiff as image type.
compression artefacts hurt my feelings.
* Cleanup
* Making Models as plugins
* Do not reload model on each image #39 + Adding FaceFilter #53
* Adding @lukaville PR for #43 and #44 (possibly)
* Training done in a separate thread
* Better log for plugin load
* Adding a prefetch to train.py #49
(Note that we prefetch 2 batches of images, due to the queue behavior)
+ More compact logging with verbose info included
* correction of DirectoryProcessor signature
* adding missing import
* Convert with parallel preprocessing of files
* Added coverage var for trainer
Added a var with comment. Feel free to add it as argument
* corrections
* Modifying preview and normalization of image + correction
* Cleanup
* Improvement of the train action usability.
* Prints the current iteration number in verbose mode.
* The number of iterations before saving the data can be changed by a command line optin.
* Added option to write training result to file even when in preview.
* Prints time elapsed for each iteration in verbose mode when training.
* Created a single script to call the other ones.
Usage is ./faceswap.py {train|extract|convert}
* Improved the help from the commands.
* Added forgotten faceswap.py file.
* Changed gitignore to add the scripts.
* Updates gitignore.
* Added guarding not to execute code when imported.
* Removed useless script. Display help when no arguments are provided.
* Update README
