mirror of
https://github.com/deepfakes/faceswap
synced 2025-06-07 10:37:19 -04:00
* 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
214 lines
7.7 KiB
Python
214 lines
7.7 KiB
Python
#!/usr/bin python3
|
|
""" Utilities available across all scripts """
|
|
|
|
import logging
|
|
import os
|
|
import warnings
|
|
|
|
from hashlib import sha1
|
|
from pathlib import Path
|
|
from re import finditer
|
|
|
|
import cv2
|
|
import numpy as np
|
|
|
|
import dlib
|
|
|
|
from lib.faces_detect import DetectedFace
|
|
from lib.logger import get_loglevel
|
|
|
|
|
|
logger = logging.getLogger(__name__) # pylint: disable=invalid-name
|
|
|
|
# Global variables
|
|
_image_extensions = [ # pylint: disable=invalid-name
|
|
".bmp", ".jpeg", ".jpg", ".png", ".tif", ".tiff"]
|
|
_video_extensions = [ # pylint: disable=invalid-name
|
|
".avi", ".flv", ".mkv", ".mov", ".mp4", ".mpeg", ".webm"]
|
|
|
|
|
|
def get_folder(path):
|
|
""" Return a path to a folder, creating it if it doesn't exist """
|
|
logger.debug("Requested path: '%s'", path)
|
|
output_dir = Path(path)
|
|
output_dir.mkdir(parents=True, exist_ok=True)
|
|
logger.debug("Returning: '%s'", output_dir)
|
|
return output_dir
|
|
|
|
|
|
def get_image_paths(directory):
|
|
""" Return a list of images that reside in a folder """
|
|
image_extensions = _image_extensions
|
|
dir_contents = list()
|
|
|
|
if not os.path.exists(directory):
|
|
logger.debug("Creating folder: '%s'", directory)
|
|
directory = get_folder(directory)
|
|
|
|
dir_scanned = sorted(os.scandir(directory), key=lambda x: x.name)
|
|
logger.debug("Scanned Folder contains %s files", len(dir_scanned))
|
|
logger.trace("Scanned Folder Contents: %s", dir_scanned)
|
|
|
|
for chkfile in dir_scanned:
|
|
if any([chkfile.name.lower().endswith(ext)
|
|
for ext in image_extensions]):
|
|
logger.trace("Adding '%s' to image list", chkfile.path)
|
|
dir_contents.append(chkfile.path)
|
|
|
|
logger.debug("Returning %s images", len(dir_contents))
|
|
return dir_contents
|
|
|
|
|
|
def hash_image_file(filename):
|
|
""" Return an image file's sha1 hash """
|
|
img = cv2.imread(filename) # pylint: disable=no-member
|
|
img_hash = sha1(img).hexdigest()
|
|
logger.trace("filename: '%s', hash: %s", filename, img_hash)
|
|
return img_hash
|
|
|
|
|
|
def hash_encode_image(image, extension):
|
|
""" Encode the image, get the hash and return the hash with
|
|
encoded image """
|
|
img = cv2.imencode(extension, image)[1] # pylint: disable=no-member
|
|
f_hash = sha1(
|
|
cv2.imdecode(img, cv2.IMREAD_UNCHANGED)).hexdigest() # pylint: disable=no-member
|
|
return f_hash, img
|
|
|
|
|
|
def backup_file(directory, filename):
|
|
""" Backup a given file by appending .bk to the end """
|
|
logger.trace("Backing up: '%s'", filename)
|
|
origfile = os.path.join(directory, filename)
|
|
backupfile = origfile + '.bk'
|
|
if os.path.exists(backupfile):
|
|
logger.trace("Removing existing file: '%s'", backup_file)
|
|
os.remove(backupfile)
|
|
if os.path.exists(origfile):
|
|
logger.trace("Renaming: '%s' to '%s'", origfile, backup_file)
|
|
os.rename(origfile, backupfile)
|
|
|
|
|
|
def set_system_verbosity(loglevel):
|
|
""" Set the verbosity level of tensorflow and suppresses
|
|
future and deprecation warnings from any modules
|
|
From:
|
|
https://stackoverflow.com/questions/35911252/disable-tensorflow-debugging-information
|
|
Can be set to:
|
|
0 - all logs shown
|
|
1 - filter out INFO logs
|
|
2 - filter out WARNING logs
|
|
3 - filter out ERROR logs """
|
|
|
|
numeric_level = get_loglevel(loglevel)
|
|
loglevel = "2" if numeric_level > 15 else "0"
|
|
logger.debug("System Verbosity level: %s", loglevel)
|
|
os.environ['TF_CPP_MIN_LOG_LEVEL'] = loglevel
|
|
if loglevel != '0':
|
|
for warncat in (FutureWarning, DeprecationWarning, UserWarning):
|
|
warnings.simplefilter(action='ignore', category=warncat)
|
|
|
|
|
|
def rotate_landmarks(face, rotation_matrix):
|
|
# pylint: disable=c-extension-no-member
|
|
""" Rotate the landmarks and bounding box for faces
|
|
found in rotated images.
|
|
Pass in a DetectedFace object, Alignments dict or DLib rectangle"""
|
|
logger.trace("Rotating landmarks: (rotation_matrix: %s, type(face): %s",
|
|
rotation_matrix, type(face))
|
|
if isinstance(face, DetectedFace):
|
|
bounding_box = [[face.x, face.y],
|
|
[face.x + face.w, face.y],
|
|
[face.x + face.w, face.y + face.h],
|
|
[face.x, face.y + face.h]]
|
|
landmarks = face.landmarksXY
|
|
|
|
elif isinstance(face, dict):
|
|
bounding_box = [[face.get("x", 0), face.get("y", 0)],
|
|
[face.get("x", 0) + face.get("w", 0),
|
|
face.get("y", 0)],
|
|
[face.get("x", 0) + face.get("w", 0),
|
|
face.get("y", 0) + face.get("h", 0)],
|
|
[face.get("x", 0),
|
|
face.get("y", 0) + face.get("h", 0)]]
|
|
landmarks = face.get("landmarksXY", list())
|
|
|
|
elif isinstance(face,
|
|
dlib.rectangle): # pylint: disable=c-extension-no-member
|
|
bounding_box = [[face.left(), face.top()],
|
|
[face.right(), face.top()],
|
|
[face.right(), face.bottom()],
|
|
[face.left(), face.bottom()]]
|
|
landmarks = list()
|
|
else:
|
|
raise ValueError("Unsupported face type")
|
|
|
|
logger.trace("Original landmarks: %s", landmarks)
|
|
|
|
rotation_matrix = cv2.invertAffineTransform( # pylint: disable=no-member
|
|
rotation_matrix)
|
|
rotated = list()
|
|
for item in (bounding_box, landmarks):
|
|
if not item:
|
|
continue
|
|
points = np.array(item, np.int32)
|
|
points = np.expand_dims(points, axis=0)
|
|
transformed = cv2.transform(points, # pylint: disable=no-member
|
|
rotation_matrix).astype(np.int32)
|
|
rotated.append(transformed.squeeze())
|
|
|
|
# Bounding box should follow x, y planes, so get min/max
|
|
# for non-90 degree rotations
|
|
pt_x = min([pnt[0] for pnt in rotated[0]])
|
|
pt_y = min([pnt[1] for pnt in rotated[0]])
|
|
pt_x1 = max([pnt[0] for pnt in rotated[0]])
|
|
pt_y1 = max([pnt[1] for pnt in rotated[0]])
|
|
|
|
if isinstance(face, DetectedFace):
|
|
face.x = int(pt_x)
|
|
face.y = int(pt_y)
|
|
face.w = int(pt_x1 - pt_x)
|
|
face.h = int(pt_y1 - pt_y)
|
|
face.r = 0
|
|
if len(rotated) > 1:
|
|
rotated_landmarks = [tuple(point) for point in rotated[1].tolist()]
|
|
face.landmarksXY = rotated_landmarks
|
|
elif isinstance(face, dict):
|
|
face["x"] = int(pt_x)
|
|
face["y"] = int(pt_y)
|
|
face["w"] = int(pt_x1 - pt_x)
|
|
face["h"] = int(pt_y1 - pt_y)
|
|
face["r"] = 0
|
|
if len(rotated) > 1:
|
|
rotated_landmarks = [tuple(point) for point in rotated[1].tolist()]
|
|
face["landmarksXY"] = rotated_landmarks
|
|
else:
|
|
rotated_landmarks = dlib.rectangle( # pylint: disable=c-extension-no-member
|
|
int(pt_x), int(pt_y), int(pt_x1), int(pt_y1))
|
|
face = rotated_landmarks
|
|
|
|
logger.trace("Rotated landmarks: %s", rotated_landmarks)
|
|
return face
|
|
|
|
|
|
def camel_case_split(identifier):
|
|
""" Split a camel case name
|
|
from: https://stackoverflow.com/questions/29916065 """
|
|
matches = finditer(
|
|
".+?(?:(?<=[a-z])(?=[A-Z])|(?<=[A-Z])(?=[A-Z][a-z])|$)",
|
|
identifier)
|
|
return [m.group(0) for m in matches]
|
|
|
|
|
|
def safe_shutdown():
|
|
""" Close queues, threads and processes in event of crash """
|
|
logger.debug("Safely shutting down")
|
|
from lib.queue_manager import queue_manager
|
|
from lib.multithreading import terminate_processes
|
|
queue_manager.terminate_queues()
|
|
terminate_processes()
|
|
logger.debug("Cleanup complete. Shutting down queue manager and exiting")
|
|
queue_manager._log_queue.put(None) # pylint: disable=protected-access
|
|
while not queue_manager._log_queue.empty(): # pylint: disable=protected-access
|
|
continue
|
|
queue_manager.manager.shutdown()
|