IMAGEIO_FFMPEG_EXE is now set

2022-09-24 22:23:59 +02:00
14 changed files with 35 additions and 288 deletions
--- a/.gitignore
+++ b/.gitignore
@@ -139,6 +139,4 @@ checkpoints/

 *.pth
 *.onnx
-wandb/
-temp_results/
-output/*.*
+wandb/
--- a/README.md
+++ b/README.md
@@ -4,8 +4,6 @@

 **Our method can realize **arbitrary face swapping** on images and videos with **one single trained model**.**

-***We are recruiting full-time engineers. If you are interested, please send an [email](mailto:chen19910528@sjtu.edu.cn?subject=[GitHub]%20Source%20Han%20Sans) to my team. Please refer to the website for specific recruitment conditions: [Requirements](https://join.sjtu.edu.cn/Admin/QsPreview.aspx?qsid=44f5413a90974114b8f5e643177ef32d)***
-
 Training and test code are now available!
 [ <a href="https://colab.research.google.com/github/neuralchen/SimSwap/blob/main/train.ipynb"><img src="https://colab.research.google.com/assets/colab-badge.svg" alt="google colab logo"></a>](https://colab.research.google.com/github/neuralchen/SimSwap/blob/main/train.ipynb)

@@ -15,12 +13,7 @@ The high resolution version of ***SimSwap-HQ*** is supported!

 [![simswaplogo](/docs/img/logo1.png)](https://github.com/neuralchen/SimSwap)

-Our paper can be downloaded from [[Arxiv]](https://arxiv.org/pdf/2106.06340v1.pdf) [[ACM DOI]](https://dl.acm.org/doi/10.1145/3394171.3413630)
-
-
-### This project also received support from [SocialBook](https://socialbook.io).
-<!-- [![logo](./simswaplogo/socialbook_logo.2020.357eed90add7705e54a8.svg)](https://socialbook.io) -->
-<img width=30% src="./simswaplogo/socialbook_logo.2020.357eed90add7705e54a8.svg"/>
+Our paper can be downloaded from [[Arxiv]](https://arxiv.org/pdf/2106.06340v1.pdf) [[ACM DOI]](https://dl.acm.org/doi/10.1145/3394171.3413630) 

 <!-- [[Google Drive]](https://drive.google.com/file/d/1fcfWOGt1mkBo7F0gXVKitf8GJMAXQxZD/view?usp=sharing) 
 [[Baidu Drive ]](https://pan.baidu.com/s/1-TKFuycRNUKut8hn4IimvA) Password: ```ummt``` -->
@@ -36,11 +29,7 @@ If you find this project useful, please star it. It is the greatest appreciation

 ## Top News <img width=8% src="./docs/img/new.gif"/>

-**`2023-09-26`**: We fixed bugs in colab!
-
-**`2023-04-25`**: We fixed the "AttributeError: 'SGD' object has no attribute 'defaults' now" bug. If you have already downloaded **arcface_checkpoint.tar**, please **download it again**. Also, you also need to update the scripts in ```./models/```.
-
-**`2022-04-21`**: For resource limited users, we provide the cropped VGGFace2-224 dataset [[Google Driver] VGGFace2-224 (10.8G)](https://drive.google.com/file/d/19pWvdEHS-CEG6tW3PdxdtZ5QEymVjImc/view?usp=sharing) [[Baidu Driver]](https://pan.baidu.com/s/1OiwLJHVBSYB4AY2vEcfN0A) [Password: lrod].
+**`2022-04-21`**: For resource limited users, we provide the cropped VGGFace2-224 dataset [[Google Driver] VGGFace2-224 (10.8G)](https://drive.google.com/file/d/19pWvdEHS-CEG6tW3PdxdtZ5QEymVjImc/view?usp=sharing) [[Baidu Driver] ](https://pan.baidu.com/s/1OiwLJHVBSYB4AY2vEcfN0A) [Password: lrod].

 **`2022-04-20`**: Training scripts are now available. We highly recommend that you guys train the simswap model with our released high quality dataset [VGGFace2-HQ](https://github.com/NNNNAI/VGGFace2-HQ).

@@ -61,9 +50,6 @@ If you find this project useful, please star it. It is the greatest appreciation

 [![logo](./docs/img/vggface2_hq_compare.png)](https://github.com/NNNNAI/VGGFace2-HQ)

-
-
-
 ## Dependencies
 - python3.6+
 - pytorch1.5+
@@ -74,7 +60,6 @@ If you find this project useful, please star it. It is the greatest appreciation
 - imageio
 - moviepy
 - insightface
- ***timm==0.5.4***

 ## Training

@@ -97,7 +82,7 @@ For faster convergence and better results, a large batch size (more than 16) is
 python train.py --name simswap224_test --batchSize 8  --gpu_ids 0 --dataset /path/to/VGGFace2HQ --Gdeep False
 ```

-[Colab demo for training 224 model][ <a href="https://colab.research.google.com/github/neuralchen/SimSwap/blob/main/train.ipynb"><img src="https://colab.research.google.com/assets/colab-badge.svg" alt="google colab logo"></a>](https://colab.research.google.com/github/neuralchen/SimSwap/blob/main/train.ipynb)
+[Colab demo for training 224 model](https://colab.research.google.com/github/neuralchen/SimSwap/blob/main/train.ipynb) 

 For faster convergence and better results, a large batch size (more than 16) is recommended!

@@ -117,7 +102,7 @@ python train.py --name simswap512_test  --batchSize 16  --gpu_ids 0 --dataset /p

 <div style="background: yellow; width:140px; font-weight:bold;font-family: sans-serif;">Stronger feature</div>

-[Colab for switching specific faces in multi-face videos][ <a href="https://colab.research.google.com/github/neuralchen/SimSwap/blob/main/MultiSpecific.ipynb"><img src="https://colab.research.google.com/assets/colab-badge.svg" alt="google colab logo"></a>](https://colab.research.google.com/github/neuralchen/SimSwap/blob/main/MultiSpecific.ipynb)
+[Colab for switching specific faces in multi-face videos](https://colab.research.google.com/github/neuralchen/SimSwap/blob/main/MultiSpecific.ipynb) 

 [Image face swapping demo & Docker image on Replicate](https://replicate.ai/neuralchen/simswap-image)

@@ -171,7 +156,7 @@ At the same time, if you have suggestions for our project, please feel free to a
 For academic and non-commercial use only.The whole project is under the CC-BY-NC 4.0 license. See [LICENSE](https://github.com/neuralchen/SimSwap/blob/main/LICENSE) for additional details.


-## To cite our papers
+## To cite our paper
 ```
@inproceedings{DBLP:conf/mm/ChenCNG20,
  author    = {Renwang Chen and
@@ -183,22 +168,6 @@ For academic and non-commercial use only.The whole project is under the CC-BY-NC
  year      = {2020}
 }
 ```
-```
-@Article{simswapplusplus,
-    author  = {Xuanhong Chen and
-              Bingbing Ni and
-              Yutian Liu and
-              Naiyuan Liu and
-              Zhilin Zeng and
-              Hang Wang},
-    title   = {SimSwap++: Towards Faster and High-Quality Identity Swapping},
-    journal = {{IEEE} Trans. Pattern Anal. Mach. Intell.},
-    volume  = {46},
-    number  = {1},
-    pages   = {576--592},
-    year    = {2024}
-}
-```

 ## Related Projects

--- a/models/init.py
+++ b/models/init.py
@@ -1,4 +1,4 @@
-from .arcface_models import ArcMarginModel
-from .arcface_models import ResNet
-from .arcface_models import IRBlock
-from .arcface_models import SEBlock
+from .models import ArcMarginModel
+from .models import ResNet
+from .models import IRBlock
+from .models import SEBlock
--- a/models/arcface_models.py
+++ b/models/arcface_models.py
@@ -1,163 +0,0 @@
-import math
-import torch
-import torch.nn.functional as F
-from torch import nn
-from torch.nn import Parameter
-from .config import device, num_classes
-
-
-
-class SEBlock(nn.Module):
-    def __init__(self, channel, reduction=16):
-        super(SEBlock, self).__init__()
-        self.avg_pool = nn.AdaptiveAvgPool2d(1)
-        self.fc = nn.Sequential(
-            nn.Linear(channel, channel // reduction),
-            nn.PReLU(),
-            nn.Linear(channel // reduction, channel),
-            nn.Sigmoid()
-        )
-
-    def forward(self, x):
-        b, c, _, _ = x.size()
-        y = self.avg_pool(x).view(b, c)
-        y = self.fc(y).view(b, c, 1, 1)
-        return x * y
-
-
-class IRBlock(nn.Module):
-    expansion = 1
-
-    def __init__(self, inplanes, planes, stride=1, downsample=None, use_se=True):
-        super(IRBlock, self).__init__()
-        self.bn0 = nn.BatchNorm2d(inplanes)
-        self.conv1 = conv3x3(inplanes, inplanes)
-        self.bn1 = nn.BatchNorm2d(inplanes)
-        self.prelu = nn.PReLU()
-        self.conv2 = conv3x3(inplanes, planes, stride)
-        self.bn2 = nn.BatchNorm2d(planes)
-        self.downsample = downsample
-        self.stride = stride
-        self.use_se = use_se
-        if self.use_se:
-            self.se = SEBlock(planes)
-
-    def forward(self, x):
-        residual = x
-        out = self.bn0(x)
-        out = self.conv1(out)
-        out = self.bn1(out)
-        out = self.prelu(out)
-
-        out = self.conv2(out)
-        out = self.bn2(out)
-        if self.use_se:
-            out = self.se(out)
-
-        if self.downsample is not None:
-            residual = self.downsample(x)
-
-        out += residual
-        out = self.prelu(out)
-
-        return out
-
-
-class ResNet(nn.Module):
-
-    def __init__(self, block, layers, use_se=True):
-        self.inplanes = 64
-        self.use_se = use_se
-        super(ResNet, self).__init__()
-        self.conv1 = nn.Conv2d(3, 64, kernel_size=3, stride=1, bias=False)
-        self.bn1 = nn.BatchNorm2d(64)
-        self.prelu = nn.PReLU()
-        self.maxpool = nn.MaxPool2d(kernel_size=2, stride=2)
-        self.layer1 = self._make_layer(block, 64, layers[0])
-        self.layer2 = self._make_layer(block, 128, layers[1], stride=2)
-        self.layer3 = self._make_layer(block, 256, layers[2], stride=2)
-        self.layer4 = self._make_layer(block, 512, layers[3], stride=2)
-        self.bn2 = nn.BatchNorm2d(512)
-        self.dropout = nn.Dropout()
-        self.fc = nn.Linear(512 * 7 * 7, 512)
-        self.bn3 = nn.BatchNorm1d(512)
-
-        for m in self.modules():
-            if isinstance(m, nn.Conv2d):
-                nn.init.xavier_normal_(m.weight)
-            elif isinstance(m, nn.BatchNorm2d) or isinstance(m, nn.BatchNorm1d):
-                nn.init.constant_(m.weight, 1)
-                nn.init.constant_(m.bias, 0)
-            elif isinstance(m, nn.Linear):
-                nn.init.xavier_normal_(m.weight)
-                nn.init.constant_(m.bias, 0)
-
-    def _make_layer(self, block, planes, blocks, stride=1):
-        downsample = None
-        if stride != 1 or self.inplanes != planes * block.expansion:
-            downsample = nn.Sequential(
-                nn.Conv2d(self.inplanes, planes * block.expansion,
-                          kernel_size=1, stride=stride, bias=False),
-                nn.BatchNorm2d(planes * block.expansion),
-            )
-
-        layers = []
-        layers.append(block(self.inplanes, planes, stride, downsample, use_se=self.use_se))
-        self.inplanes = planes
-        for i in range(1, blocks):
-            layers.append(block(self.inplanes, planes, use_se=self.use_se))
-
-        return nn.Sequential(*layers)
-
-    def forward(self, x):
-        x = self.conv1(x)
-        x = self.bn1(x)
-        x = self.prelu(x)
-        x = self.maxpool(x)
-
-        x = self.layer1(x)
-        x = self.layer2(x)
-        x = self.layer3(x)
-        x = self.layer4(x)
-
-        x = self.bn2(x)
-        x = self.dropout(x)
-        # feature = x
-        x = x.view(x.size(0), -1)
-        x = self.fc(x)
-        x = self.bn3(x)
-
-        return x
-
-
-class ArcMarginModel(nn.Module):
-    def __init__(self, args):
-        super(ArcMarginModel, self).__init__()
-
-        self.weight = Parameter(torch.FloatTensor(num_classes, args.emb_size))
-        nn.init.xavier_uniform_(self.weight)
-
-        self.easy_margin = args.easy_margin
-        self.m = args.margin_m
-        self.s = args.margin_s
-
-        self.cos_m = math.cos(self.m)
-        self.sin_m = math.sin(self.m)
-        self.th = math.cos(math.pi - self.m)
-        self.mm = math.sin(math.pi - self.m) * self.m
-
-    def forward(self, input, label):
-        x = F.normalize(input)
-        W = F.normalize(self.weight)
-        cosine = F.linear(x, W)
-        sine = torch.sqrt(1.0 - torch.pow(cosine, 2))
-        phi = cosine * self.cos_m - sine * self.sin_m  # cos(theta + m)
-        if self.easy_margin:
-            phi = torch.where(cosine > 0, phi, cosine)
-        else:
-            phi = torch.where(cosine > self.th, phi, cosine - self.mm)
-        one_hot = torch.zeros(cosine.size(), device=device)
-        one_hot.scatter_(1, label.view(-1, 1).long(), 1)
-        output = (one_hot * phi) + ((1.0 - one_hot) * cosine)
-        output *= self.s
-        return output
--- a/models/fs_model.py
+++ b/models/fs_model.py
@@ -59,10 +59,13 @@ class fsModel(BaseModel):
        self.netG = Generator_Adain_Upsample(input_nc=3, output_nc=3, latent_size=512, n_blocks=9, deep=False)
        self.netG.to(device)

+
+
+
        # Id network
        netArc_checkpoint = opt.Arc_path
-        netArc_checkpoint = torch.load(netArc_checkpoint, map_location=torch.device("cpu"))
-        self.netArc = netArc_checkpoint
+        netArc_checkpoint = torch.load(netArc_checkpoint)
+        self.netArc = netArc_checkpoint['model'].module
        self.netArc = self.netArc.to(device)
        self.netArc.eval()

--- a/models/models.py
+++ b/models/models.py
@@ -7,9 +7,13 @@ from .config import device, num_classes


 def create_model(opt):
-    #from .pix2pixHD_model import Pix2PixHDModel, InferenceModel
-    from .fs_model import fsModel
-    model = fsModel()
+    if opt.model == 'pix2pixHD':
+        #from .pix2pixHD_model import Pix2PixHDModel, InferenceModel
+        from .fs_model import fsModel
+        model = fsModel()
+    else:
+        from .ui_model import UIModel
+        model = UIModel()

    model.initialize(opt)
    if opt.verbose:
--- a/models/projected_model.py
+++ b/models/projected_model.py
@@ -5,7 +5,7 @@
 # Created Date: Wednesday January 12th 2022
 # Author: Chen Xuanhong
 # Email: chenxuanhongzju@outlook.com
-# Last Modified:  Saturday, 13th May 2023 9:56:35 am
+# Last Modified:  Thursday, 21st April 2022 8:13:37 pm
 # Modified By: Chen Xuanhong
 # Copyright (c) 2022 Shanghai Jiao Tong University
 #############################################################
@@ -46,7 +46,7 @@ class fsModel(BaseModel):
        # Id network
        netArc_checkpoint = opt.Arc_path
        netArc_checkpoint = torch.load(netArc_checkpoint, map_location=torch.device("cpu"))
-        self.netArc = netArc_checkpoint
+        self.netArc = netArc_checkpoint['model'].module
        self.netArc = self.netArc.cuda()
        self.netArc.eval()
        self.netArc.requires_grad_(False)
--- a/options/base_options.py
+++ b/options/base_options.py
@@ -13,7 +13,7 @@ class BaseOptions():
        self.parser.add_argument('--name', type=str, default='people', help='name of the experiment. It decides where to store samples and models')
        self.parser.add_argument('--gpu_ids', type=str, default='0', help='gpu ids: e.g. 0  0,1,2, 0,2. use -1 for CPU')
        self.parser.add_argument('--checkpoints_dir', type=str, default='./checkpoints', help='models are saved here')
-        # self.parser.add_argument('--model', type=str, default='pix2pixHD', help='which model to use')
+        self.parser.add_argument('--model', type=str, default='pix2pixHD', help='which model to use')
        self.parser.add_argument('--norm', type=str, default='batch', help='instance normalization or batch normalization')
        self.parser.add_argument('--use_dropout', action='store_true', help='use dropout for the generator')
        self.parser.add_argument('--data_type', default=32, type=int, choices=[8, 16, 32], help="Supported data type i.e. 8, 16, 32 bit")
--- a/options/test_options.py
+++ b/options/test_options.py
@@ -22,17 +22,17 @@ class TestOptions(BaseOptions):
        self.parser.add_argument("--export_onnx", type=str, help="export ONNX model to a given file")
        self.parser.add_argument("--engine", type=str, help="run serialized TRT engine")
        self.parser.add_argument("--onnx", type=str, help="run ONNX model via TRT")        
-        self.parser.add_argument("--Arc_path", type=str, default='arcface_model/arcface_checkpoint.tar', help="run ONNX model via TRT")
-        self.parser.add_argument("--pic_a_path", type=str, default='G:/swap_data/ID/elon-musk-hero-image.jpeg', help="Person who provides identity information")
-        self.parser.add_argument("--pic_b_path", type=str, default='./demo_file/multi_people.jpg', help="Person who provides information other than their identity")
+        self.parser.add_argument("--Arc_path", type=str, default='models/BEST_checkpoint.tar', help="run ONNX model via TRT")
+        self.parser.add_argument("--pic_a_path", type=str, default='./crop_224/gdg.jpg', help="Person who provides identity information")
+        self.parser.add_argument("--pic_b_path", type=str, default='./crop_224/zrf.jpg', help="Person who provides information other than their identity")
        self.parser.add_argument("--pic_specific_path", type=str, default='./crop_224/zrf.jpg', help="The specific person to be swapped")
        self.parser.add_argument("--multisepcific_dir", type=str, default='./demo_file/multispecific', help="Dir for multi specific")
-        self.parser.add_argument("--video_path", type=str, default='G:/swap_data/video/HSB_Demo_Trim.mp4', help="path for the video to swap")
+        self.parser.add_argument("--video_path", type=str, default='./demo_file/multi_people_1080p.mp4', help="path for the video to swap")
        self.parser.add_argument("--temp_path", type=str, default='./temp_results', help="path to save temporarily images")
        self.parser.add_argument("--output_path", type=str, default='./output/', help="results path")
        self.parser.add_argument('--id_thres', type=float, default=0.03, help='how many test images to run')
        self.parser.add_argument('--no_simswaplogo', action='store_true', help='Remove the watermark')
        self.parser.add_argument('--use_mask', action='store_true', help='Use mask for better result')
-        self.parser.add_argument('--crop_size', type=int, default=512, help='Crop of size of input image')
+        self.parser.add_argument('--crop_size', type=int, default=224, help='Crop of size of input image')
        
        self.isTrain = False
--- a/output/result.jpg
+++ b/output/result.jpg
--- a/simswaplogo/socialbook_logo.2020.357eed90add7705e54a8.svg
+++ b/simswaplogo/socialbook_logo.2020.357eed90add7705e54a8.svg
@@ -1,64 +0,0 @@
-<?xml version="1.0" encoding="utf-8"?>
-<!-- Generator: Adobe Illustrator 24.1.0, SVG Export Plug-In . SVG Version: 6.00 Build 0)  -->
-<svg version="1.1" id="Layer_1" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" x="0px" y="0px"
-	 viewBox="0 0 663 121" style="enable-background:new 0 0 663 121;" xml:space="preserve">
-<style type="text/css">
-	.st0{fill:#3753FA;}
-	.st1{fill:#FFFFFF;}
-	.st2{fill:none;}
-</style>
-<path class="st0" d="M64.92,86.76c0-12.55-10.28-23.2-23.21-23.2h-8.76c-1.52,0.38-2.28-0.76-2.28-1.9c0-1.14,0.76-2.28,2.28-2.28
-	c0,0,0,0,0.38,0h31.96c0-11.41-9.52-20.91-20.92-20.91H32.95c-12.94,0-23.21,10.27-23.21,23.2c0,12.93,10.28,22.82,23.21,22.82h8.76
-	c1.14,0,2.28,0.76,2.28,2.28c0,1.52-0.76,2.28-2.28,2.28c0,0,0,0-0.38,0H16.59c-0.76,9.13-5.7,17.87-13.7,20.91H41.7
-	C54.26,109.96,64.92,99.69,64.92,86.76"/>
-<g>
-	<path class="st0" d="M179.91,45.59c4.33,2.47,8.04,6.18,10.51,10.51c2.47,4.33,3.71,9.9,3.71,15.46c0,6.18-1.24,11.13-3.71,16.07
-		c-2.47,4.33-6.18,8.04-10.51,10.51c-4.33,1.86-9.27,3.09-15.46,3.09c-5.57,0-10.51-1.24-14.84-3.71
-		c-4.33-2.47-8.04-6.18-10.51-10.51c-2.47-4.33-3.71-9.9-3.71-16.07c0-6.17,1.24-11.13,3.71-15.46c2.47-4.33,6.18-8.04,10.51-10.51
-		s9.27-3.71,14.84-3.71C170.63,41.88,175.58,43.12,179.91,45.59z M156.42,54.86c-2.47,1.24-4.94,3.71-6.18,6.18
-		c-1.86,3.09-2.47,6.18-2.47,10.51c0,4.33,0.62,8.04,2.47,10.51c1.86,3.09,3.71,4.94,6.18,6.18s5.57,1.86,8.66,1.86
-		s6.18-0.62,8.66-1.86s4.94-3.71,6.18-6.18c1.86-3.09,2.47-6.18,2.47-10.51c0-4.33-0.62-8.04-2.47-10.51
-		c-1.86-2.47-3.71-4.94-6.18-6.18c-2.47-1.24-5.57-1.86-8.66-1.86C161.36,52.39,158.89,53,156.42,54.86z M252.86,48.06
-		c4.94,3.71,8.04,9.27,9.27,16.07h-12.98c-0.62-3.71-2.47-6.8-4.94-8.66c-2.47-1.86-6.18-3.09-9.9-3.09
-		c-4.33,0-8.66,1.86-11.74,4.94s-4.33,8.04-4.33,14.22c0,6.18,1.24,11.13,4.33,14.22c3.09,3.1,6.8,4.94,11.74,4.94
-		c4.33,0,7.42-1.24,9.9-3.09c2.47-1.86,4.33-4.94,4.94-8.66h12.98c-1.24,6.8-4.33,12.37-9.27,16.07s-11.13,6.18-18.55,6.18
-		c-5.57,0-10.51-1.24-14.84-3.71c-4.33-2.47-8.04-6.18-9.9-10.51c-2.47-4.33-3.71-9.9-3.71-16.07c0-6.18,1.24-11.13,3.71-15.46
-		c1.86-3.71,5.57-7.42,9.9-9.9c4.33-2.47,9.27-3.71,14.84-3.71C241.73,41.88,247.92,44.35,252.86,48.06L252.86,48.06z M276.36,20.85
-		c1.24-1.24,3.09-1.86,5.57-1.86c2.47,0,4.33,0.62,5.57,1.86c1.24,1.24,1.86,3.09,1.86,4.94c0,1.86-0.62,3.71-1.86,4.94
-		c-1.24,1.24-3.09,1.86-5.57,1.86c-2.47,0-4.33-0.62-5.57-1.86c-1.24-1.24-1.86-3.09-1.86-4.94
-		C274.5,23.94,275.12,22.09,276.36,20.85z M288.11,43.12v58.12h-11.74V43.12L288.11,43.12L288.11,43.12z M341.9,46.2
-		c3.71,2.47,6.18,6.18,8.04,10.51v-13.6h11.74v58.12h-11.74v-13.6c-1.24,4.33-4.33,8.04-8.04,10.51c-3.71,2.47-8.04,3.71-13.6,3.71
-		c-4.94,0-9.27-1.24-13.6-3.71c-3.71-2.47-6.8-5.57-9.27-10.51c-2.47-4.94-3.09-9.9-3.09-16.07c0-6.18,1.24-11.13,3.09-15.46
-		c2.47-4.33,5.57-8.04,9.27-10.51s8.66-3.71,13.6-3.71C333.86,41.88,338.19,43.73,341.9,46.2L341.9,46.2z M319.63,57.95
-		c-3.09,3.09-4.94,8.04-4.94,14.22c0,6.18,1.86,10.51,4.94,14.22c3.09,3.71,7.42,4.94,12.98,4.94c3.09,0,6.18-0.62,9.27-2.47
-		c2.47-1.86,4.94-3.71,6.18-6.8c1.24-3.09,2.47-6.18,2.47-9.9c0-3.71-0.62-7.42-2.47-9.9c-1.86-2.47-3.71-4.94-6.18-6.8
-		c-2.47-1.86-5.57-2.47-9.27-2.47C327.06,52.39,322.73,54.24,319.63,57.95z M390.74,22.71v77.9h-11.74v-77.9H390.74L390.74,22.71z
-		 M457.51,69.08c2.47,3.09,4.33,7.42,4.33,11.74c0,6.18-1.86,11.13-6.18,14.84c-4.33,3.71-9.9,5.57-17.93,5.57h-29.68V27.05h29.05
-		c7.42,0,12.98,1.86,17.31,4.94c3.71,3.09,6.18,8.04,6.18,13.6c0,4.94-1.24,8.66-3.71,11.13c-2.47,2.47-5.57,4.94-9.9,5.57
-		C451.33,63.52,455.04,65.37,457.51,69.08z M419.8,58.57h14.84c4.33,0,7.42-1.24,9.9-3.09c2.47-1.86,3.09-4.94,3.09-8.04
-		c0-3.71-1.24-6.18-3.09-8.04c-2.47-1.86-5.57-3.09-9.9-3.09h-14.22L419.8,58.57L419.8,58.57z M445.77,87.63
-		c2.47-1.86,3.71-4.94,3.71-8.66s-1.24-6.8-3.71-8.66c-2.47-1.86-6.18-3.09-10.51-3.09h-14.84v23.5h14.84
-		C439.58,90.72,443.29,90.1,445.77,87.63z M517.48,45.59c4.33,2.47,8.04,5.57,10.51,10.51c2.47,4.33,3.71,9.9,3.71,15.46
-		c0,6.18-1.24,11.13-3.71,16.07c-2.47,4.33-6.18,8.04-10.51,10.51c-4.33,2.47-9.27,3.71-14.84,3.71s-10.51-1.24-14.84-3.71
-		c-4.33-2.47-8.04-6.18-10.51-10.51s-3.71-9.9-3.71-16.07c0-6.18,1.24-11.13,3.71-15.46s6.18-8.04,10.51-10.51
-		c4.33-2.47,9.27-3.71,14.84-3.71C507.59,41.88,512.54,43.12,517.48,45.59z M493.37,54.86c-2.47,1.24-4.94,3.71-6.18,6.18
-		c-1.86,3.09-2.47,6.18-2.47,10.51c0,4.33,0.62,8.04,2.47,10.51c1.86,3.09,3.71,4.94,6.18,6.18c2.47,1.24,5.57,1.86,8.66,1.86
-		s6.18-0.62,8.66-1.86c2.47-1.24,4.94-3.71,6.18-6.18c1.86-3.09,2.47-6.18,2.47-10.51c0-4.33-0.62-8.04-2.47-10.51
-		s-3.71-4.94-6.18-6.18c-2.47-1.24-5.57-1.86-8.66-1.86C498.93,52.39,495.84,53,493.37,54.86z M587.34,45.59
-		c4.33,2.47,8.04,5.57,10.51,10.51c2.47,4.33,3.71,9.9,3.71,15.46c0,6.18-1.24,11.13-3.71,16.07c-2.47,4.33-6.18,8.04-10.51,10.51
-		c-4.33,2.47-9.27,3.71-14.84,3.71c-5.57,0-10.51-1.24-14.84-3.71c-4.33-2.47-8.04-6.18-10.51-10.51s-3.71-9.9-3.71-16.07
-		c0-6.18,1.24-11.13,3.71-15.46s6.18-8.04,10.51-10.51c4.33-2.47,9.27-3.71,14.84-3.71S583.02,43.12,587.34,45.59L587.34,45.59z
-		 M563.85,54.86c-2.47,1.24-4.94,3.71-6.18,6.18c-1.86,3.09-2.47,6.18-2.47,10.51c0,4.33,0.62,8.04,2.47,10.51
-		c1.86,3.09,3.71,4.94,6.18,6.18s5.57,1.86,8.66,1.86c3.09,0,6.18-0.62,8.66-1.86s4.94-3.71,6.18-6.18
-		c1.86-3.09,2.47-6.18,2.47-10.51c0-4.33-0.62-8.04-2.47-10.51c-1.86-2.47-3.71-4.94-6.18-6.18c-2.47-1.24-5.57-1.86-8.66-1.86
-		C568.8,52.39,566.33,53,563.85,54.86z M648.56,100.61l-21.02-25.97v25.97h-11.74v-77.9h11.74v44.52l20.4-24.73h14.22l-25.35,29.05
-		l25.97,29.05L648.56,100.61L648.56,100.61z"/>
-	<path class="st0" d="M66.71,121L66.71,121c-33.3,0-60.3-27-60.3-60.3l0,0c0-33.3,27-60.3,60.3-60.3l0,0c33.3,0,60.3,27,60.3,60.3
-		l0,0C127.01,94,100.01,121,66.71,121z"/>
-</g>
-<path class="st1" d="M95.72,73.25c0-12.55-10.28-23.2-23.21-23.2h-8.76c-1.52,0.38-2.28-0.76-2.28-1.9s0.76-2.28,2.28-2.28
-	c0,0,0,0,0.38,0h31.96c0-11.41-9.52-20.91-20.92-20.91H63.75c-12.94,0-23.21,10.27-23.21,23.2s10.28,22.82,23.21,22.82h8.76
-	c1.14,0,2.28,0.76,2.28,2.28s-0.76,2.28-2.28,2.28c0,0,0,0-0.38,0H47.39c-0.76,9.13-5.7,17.87-13.7,20.91H72.5
-	C85.06,96.45,95.72,86.18,95.72,73.25"/>
-<rect id="_x3C_Slice_x3E_" x="6" class="st2" width="657" height="121"/>
-</svg>
--- a/test_one_image.py
+++ b/test_one_image.py
@@ -83,4 +83,4 @@ if __name__ == '__main__':

        output = output*255

-        cv2.imwrite(opt.output_path + 'result.jpg', output)
+        cv2.imwrite(opt.output_path + 'result.jpg',output)
--- a/train.ipynb
+++ b/train.ipynb
@@ -112,7 +112,7 @@
      "outputs": [],
      "source": [
        "!pip install googledrivedownloader\n",
-        "!pip install timm==0.5.4\n",
+        "!pip install timm\n",
        "!wget -P SimSwap/arcface_model https://github.com/neuralchen/SimSwap/releases/download/1.0/arcface_checkpoint.tar"
      ]
    },
@@ -140,10 +140,9 @@
      },
      "outputs": [],
      "source": [
-        "%mkdir /content/TrainingData\n",
        "!wget --load-cookies /tmp/cookies.txt \"https://docs.google.com/uc?export=download&confirm=$(wget --quiet --save-cookies /tmp/cookies.txt --keep-session-cookies --no-check-certificate 'https://docs.google.com/uc?export=download&id=19pWvdEHS-CEG6tW3PdxdtZ5QEymVjImc' -O- | sed -rn 's/.*confirm=([0-9A-Za-z_]+).*/\\1\\n/p')&id=19pWvdEHS-CEG6tW3PdxdtZ5QEymVjImc\" -O /content/TrainingData/vggface2_crop_arcfacealign_224.tar && rm -rf /tmp/cookies.txt\n",
        "%cd /content/\n",
-        "!tar -xzvf /content/TrainingData/vggface2_crop_arcfacealign_224.tar -C /content/TrainingData\n ",
+        "!tar -xzvf /content/TrainingData/vggface2_crop_arcfacealign_224.tar\n",
        "!rm /content/TrainingData/vggface2_crop_arcfacealign_224.tar"
      ]
    },
--- a/util/videoswap.py
+++ b/util/videoswap.py
@@ -6,7 +6,8 @@ LastEditors: Naiyuan liu
 LastEditTime: 2021-11-24 19:19:52
 Description: 
 '''
-import os 
+import os
+os.environ["IMAGEIO_FFMPEG_EXE"] = "/usr/bin/ffmpeg"
 import cv2
 import glob
 import torch