mirror of
https://github.com/iperov/DeepFaceLab.git
synced 2025-08-14 02:37:00 -07:00
New script:
5.XSeg) data_dst/src mask for XSeg trainer - fetch.bat Copies faces containing XSeg polygons to aligned_xseg\ dir. Useful only if you want to collect labeled faces and reuse them in other fakes. Now you can use trained XSeg mask in the SAEHD training process. It’s mean default ‘full_face’ mask obtained from landmarks will be replaced with the mask obtained from the trained XSeg model. use 5.XSeg.optional) trained mask for data_dst/data_src - apply.bat 5.XSeg.optional) trained mask for data_dst/data_src - remove.bat Normally you don’t need it. You can use it, if you want to use ‘face_style’ and ‘bg_style’ with obstructions. XSeg trainer : now you can choose type of face XSeg trainer : now you can restart training in “override settings” Merger: XSeg-* modes now can be used with all types of faces. Therefore old MaskEditor, FANSEG models, and FAN-x modes have been removed, because the new XSeg solution is better, simpler and more convenient, which costs only 1 hour of manual masking for regular deepfake.
This commit is contained in:
Colombo
parent
e5bad483ca
commit
6d3607a13d
30 changed files with 279 additions and 1520 deletions
|
|
@ -1,109 +0,0 @@
|
|||
import numpy as np
|
||||
import cv2
|
||||
|
||||
class IEPolysPoints:
|
||||
def __init__(self, IEPolys_parent, type):
|
||||
self.parent = IEPolys_parent
|
||||
self.type = type
|
||||
self.points = np.empty( (0,2), dtype=np.int32 )
|
||||
self.n_max = self.n = 0
|
||||
|
||||
def add(self,x,y):
|
||||
self.points = np.append(self.points[0:self.n], [ (x,y) ], axis=0)
|
||||
self.n_max = self.n = self.n + 1
|
||||
self.parent.dirty = True
|
||||
|
||||
def n_dec(self):
|
||||
self.n = max(0, self.n-1)
|
||||
self.parent.dirty = True
|
||||
return self.n
|
||||
|
||||
def n_inc(self):
|
||||
self.n = min(len(self.points), self.n+1)
|
||||
self.parent.dirty = True
|
||||
return self.n
|
||||
|
||||
def n_clip(self):
|
||||
self.points = self.points[0:self.n]
|
||||
self.n_max = self.n
|
||||
|
||||
def cur_point(self):
|
||||
return self.points[self.n-1]
|
||||
|
||||
def points_to_n(self):
|
||||
return self.points[0:self.n]
|
||||
|
||||
def set_points(self, points):
|
||||
self.points = np.array(points)
|
||||
self.n_max = self.n = len(points)
|
||||
self.parent.dirty = True
|
||||
|
||||
class IEPolys:
|
||||
def __init__(self):
|
||||
self.list = []
|
||||
self.n_max = self.n = 0
|
||||
self.dirty = True
|
||||
|
||||
def add(self, type):
|
||||
self.list = self.list[0:self.n]
|
||||
l = IEPolysPoints(self, type)
|
||||
self.list.append ( l )
|
||||
self.n_max = self.n = self.n + 1
|
||||
self.dirty = True
|
||||
return l
|
||||
|
||||
def n_dec(self):
|
||||
self.n = max(0, self.n-1)
|
||||
self.dirty = True
|
||||
return self.n
|
||||
|
||||
def n_inc(self):
|
||||
self.n = min(len(self.list), self.n+1)
|
||||
self.dirty = True
|
||||
return self.n
|
||||
|
||||
def n_list(self):
|
||||
return self.list[self.n-1]
|
||||
|
||||
def n_clip(self):
|
||||
self.list = self.list[0:self.n]
|
||||
self.n_max = self.n
|
||||
if self.n > 0:
|
||||
self.list[-1].n_clip()
|
||||
|
||||
def __iter__(self):
|
||||
for n in range(self.n):
|
||||
yield self.list[n]
|
||||
|
||||
def switch_dirty(self):
|
||||
d = self.dirty
|
||||
self.dirty = False
|
||||
return d
|
||||
|
||||
def overlay_mask(self, mask):
|
||||
h,w,c = mask.shape
|
||||
white = (1,)*c
|
||||
black = (0,)*c
|
||||
for n in range(self.n):
|
||||
poly = self.list[n]
|
||||
if poly.n > 0:
|
||||
cv2.fillPoly(mask, [poly.points_to_n()], white if poly.type == 1 else black )
|
||||
|
||||
def get_total_points(self):
|
||||
return sum([self.list[n].n for n in range(self.n)])
|
||||
|
||||
def dump(self):
|
||||
result = []
|
||||
for n in range(self.n):
|
||||
l = self.list[n]
|
||||
result += [ (l.type, l.points_to_n().tolist() ) ]
|
||||
return result
|
||||
|
||||
@staticmethod
|
||||
def load(ie_polys=None):
|
||||
obj = IEPolys()
|
||||
if ie_polys is not None and isinstance(ie_polys, list):
|
||||
for (type, points) in ie_polys:
|
||||
obj.add(type)
|
||||
obj.n_list().set_points(points)
|
||||
return obj
|
||||
|
|
@ -15,7 +15,6 @@ from .color_transfer import color_transfer, color_transfer_mix, color_transfer_s
|
|||
|
||||
from .common import normalize_channels, cut_odd_image, overlay_alpha_image
|
||||
|
||||
from .IEPolys import IEPolys
|
||||
from .SegIEPolys import *
|
||||
|
||||
from .blursharpen import LinearMotionBlur, blursharpen
|
||||
|
|
|
|||
|
|
@ -1,92 +0,0 @@
|
|||
"""
|
||||
using https://github.com/ternaus/TernausNet
|
||||
TernausNet: U-Net with VGG11 Encoder Pre-Trained on ImageNet for Image Segmentation
|
||||
"""
|
||||
|
||||
from core.leras import nn
|
||||
tf = nn.tf
|
||||
|
||||
class Ternaus(nn.ModelBase):
|
||||
def on_build(self, in_ch, base_ch):
|
||||
|
||||
self.features_0 = nn.Conv2D (in_ch, base_ch, kernel_size=3, padding='SAME')
|
||||
self.features_3 = nn.Conv2D (base_ch, base_ch*2, kernel_size=3, padding='SAME')
|
||||
self.features_6 = nn.Conv2D (base_ch*2, base_ch*4, kernel_size=3, padding='SAME')
|
||||
self.features_8 = nn.Conv2D (base_ch*4, base_ch*4, kernel_size=3, padding='SAME')
|
||||
self.features_11 = nn.Conv2D (base_ch*4, base_ch*8, kernel_size=3, padding='SAME')
|
||||
self.features_13 = nn.Conv2D (base_ch*8, base_ch*8, kernel_size=3, padding='SAME')
|
||||
self.features_16 = nn.Conv2D (base_ch*8, base_ch*8, kernel_size=3, padding='SAME')
|
||||
self.features_18 = nn.Conv2D (base_ch*8, base_ch*8, kernel_size=3, padding='SAME')
|
||||
|
||||
self.blurpool_0 = nn.BlurPool (filt_size=3)
|
||||
self.blurpool_3 = nn.BlurPool (filt_size=3)
|
||||
self.blurpool_8 = nn.BlurPool (filt_size=3)
|
||||
self.blurpool_13 = nn.BlurPool (filt_size=3)
|
||||
self.blurpool_18 = nn.BlurPool (filt_size=3)
|
||||
|
||||
self.conv_center = nn.Conv2D (base_ch*8, base_ch*8, kernel_size=3, padding='SAME')
|
||||
|
||||
self.conv1_up = nn.Conv2DTranspose (base_ch*8, base_ch*4, kernel_size=3, padding='SAME')
|
||||
self.conv1 = nn.Conv2D (base_ch*12, base_ch*8, kernel_size=3, padding='SAME')
|
||||
|
||||
self.conv2_up = nn.Conv2DTranspose (base_ch*8, base_ch*4, kernel_size=3, padding='SAME')
|
||||
self.conv2 = nn.Conv2D (base_ch*12, base_ch*8, kernel_size=3, padding='SAME')
|
||||
|
||||
self.conv3_up = nn.Conv2DTranspose (base_ch*8, base_ch*2, kernel_size=3, padding='SAME')
|
||||
self.conv3 = nn.Conv2D (base_ch*6, base_ch*4, kernel_size=3, padding='SAME')
|
||||
|
||||
self.conv4_up = nn.Conv2DTranspose (base_ch*4, base_ch, kernel_size=3, padding='SAME')
|
||||
self.conv4 = nn.Conv2D (base_ch*3, base_ch*2, kernel_size=3, padding='SAME')
|
||||
|
||||
self.conv5_up = nn.Conv2DTranspose (base_ch*2, base_ch//2, kernel_size=3, padding='SAME')
|
||||
self.conv5 = nn.Conv2D (base_ch//2+base_ch, base_ch, kernel_size=3, padding='SAME')
|
||||
|
||||
self.out_conv = nn.Conv2D (base_ch, 1, kernel_size=3, padding='SAME')
|
||||
|
||||
def forward(self, inp):
|
||||
x, = inp
|
||||
|
||||
x = x0 = tf.nn.relu(self.features_0(x))
|
||||
x = self.blurpool_0(x)
|
||||
|
||||
x = x1 = tf.nn.relu(self.features_3(x))
|
||||
x = self.blurpool_3(x)
|
||||
|
||||
x = tf.nn.relu(self.features_6(x))
|
||||
x = x2 = tf.nn.relu(self.features_8(x))
|
||||
x = self.blurpool_8(x)
|
||||
|
||||
x = tf.nn.relu(self.features_11(x))
|
||||
x = x3 = tf.nn.relu(self.features_13(x))
|
||||
x = self.blurpool_13(x)
|
||||
|
||||
x = tf.nn.relu(self.features_16(x))
|
||||
x = x4 = tf.nn.relu(self.features_18(x))
|
||||
x = self.blurpool_18(x)
|
||||
|
||||
x = self.conv_center(x)
|
||||
|
||||
x = tf.nn.relu(self.conv1_up(x))
|
||||
x = tf.concat( [x,x4], nn.conv2d_ch_axis)
|
||||
x = tf.nn.relu(self.conv1(x))
|
||||
|
||||
x = tf.nn.relu(self.conv2_up(x))
|
||||
x = tf.concat( [x,x3], nn.conv2d_ch_axis)
|
||||
x = tf.nn.relu(self.conv2(x))
|
||||
|
||||
x = tf.nn.relu(self.conv3_up(x))
|
||||
x = tf.concat( [x,x2], nn.conv2d_ch_axis)
|
||||
x = tf.nn.relu(self.conv3(x))
|
||||
|
||||
x = tf.nn.relu(self.conv4_up(x))
|
||||
x = tf.concat( [x,x1], nn.conv2d_ch_axis)
|
||||
x = tf.nn.relu(self.conv4(x))
|
||||
|
||||
x = tf.nn.relu(self.conv5_up(x))
|
||||
x = tf.concat( [x,x0], nn.conv2d_ch_axis)
|
||||
x = tf.nn.relu(self.conv5(x))
|
||||
|
||||
logits = self.out_conv(x)
|
||||
return logits, tf.nn.sigmoid(logits)
|
||||
|
||||
nn.Ternaus = Ternaus
|
||||
|
|
@ -1,5 +1,4 @@
|
|||
from .ModelBase import *
|
||||
from .PatchDiscriminator import *
|
||||
from .CodeDiscriminator import *
|
||||
from .Ternaus import *
|
||||
from .XSeg import *
|
||||
Loading…
Add table
Add a link
Reference in a new issue