TypeError: __init__() takes 3 positional arguments but 7 were given

[justin4ai]

I tried python inference.py --config configs/inference.yaml --audio_path=audio/hama_8.wav --image_source=image/justin.jpeg --cross_id --if_extract --output_dir=./docs/demo/output --inversion_option=optimize for generating talking faces with driving audio.

But the terminal raises the following error:

Cross-id testing
Load pre-trained e4e Encoder from checkpoints/Encoder_e4e.pth done.
Load pre-trained hfgi encoder from checkpoints/hfgi.pth done.
Load pre-trained StyleGAN2 from checkpoints/StyleGAN_e4e.pth done.
Stage: inference
Load pre-trained StyleHEAT [net_G_ema] from checkpoints/StyleHEAT_visual.pt done
----------------- Options ---------------
                add_image: True                          
               bfm_folder: checkpoints/BFM                      [default: BFM]
                bfm_model: BFM_model_front.mat           
                 camera_d: 10.0                          
                   center: 112.0                         
          checkpoints_dir: checkpoints                          [default: ./checkpoints]
             dataset_mode: None                          
                 ddp_port: 12355                         
        display_per_batch: True                          
                    epoch: 20                                   [default: latest]
          eval_batch_nums: inf                           
                    focal: 1015.0                        
                  gpu_ids: 0                             
               img_folder: temp                                 [default: examples]
                init_path: checkpoints/init_model/resnet50-0676ba61.pth
                  isTrain: False                                [default: None]
                    model: facerecon                     
                     name: Deep3D                               [default: face_recon]
                net_recon: resnet50                      
                    phase: test                          
                   suffix:                               
                  use_ddp: False                                [default: True]
              use_last_fc: False                         
                  verbose: False                         
           vis_batch_nums: 1                             
               world_size: 1                             
                    z_far: 15.0                          
                   z_near: 5.0                           
----------------- End -------------------
model [FaceReconModel] was created
loading the model from checkpoints/Deep3D/epoch_20.pth
  0%|                                                                                                                                     | 0/1 [00:15<?, ?it/s]
Traceback (most recent call last):
  File "inference.py", line 340, in <module>
    main()
  File "inference.py", line 327, in main
    audio_reenactment(generator, data, args.audio_path)
  File "inference.py", line 112, in audio_reenactment
    wav2lip_checkpoint, device)
TypeError: __init__() takes 3 positional arguments but 7 were given

Tracking back the error messages, I realized that, in line 111 of inference.py, Audio2Coeff class is taking 6 arguments while the constructor of Audio2Coeff (line 27 in test_audio2coeff.py - path : ./third_part/SadTalker/src) is taking only 3 arguments.

Other than the inconsistent number of arguments, I don't understand why the error message says '7 were given', not '6 were given'.

�One more thing I realized is that even the constructor doesn't take 3 arguments. Actually it takes only two excluding self argument. Any explanation?

I'll be so glad if you help me with this.
Thank you for providing such a great work!

[justin4ai]

In addition, I just checked variables for paths. They look invalid in my opinion.

[Mayy1994]

You can use the old release version of SadTalker v0.0.1, and change the checkpoint_dir = 'checkpoints/'

[justin4ai]

You can use the old release version of SadTalker v0.0.1, and change the checkpoint_dir = 'checkpoints/'

Thank you so much, but seemingly SadTalker v0.0.2 and v0.0.1 have different directory structure. Which version's structure should I follow? Because just placing SadTalker v0.0.1 (after renaming to SadTalker) under third_part doesn't work because of the directory structure.

[Mayy1994]

Here is my folder architecture. I created a new directory "src" and copied some files into it. You can also delete duplicate files.
The "utils" is copied from the new SadTalker.

Also, change line 9-11 of "test_audio2coeff.py" to
from src.audio2pose_models.audio2pose import Audio2Pose
from src.audio2exp_models.networks import SimpleWrapperV2
from src.audio2exp_models.audio2exp import Audio2Exp

I can only remember these changes I made. You may also need to dubug in some other places, but not much.

[justin4ai]

Here is my folder architecture. I created a new directory "src" and copied some files into it. You can also delete duplicate files. The "utils" is copied from the new SadTalker.

Also, change line 9-11 of "test_audio2coeff.py" to from src.audio2pose_models.audio2pose import Audio2Pose from src.audio2exp_models.networks import SimpleWrapperV2 from src.audio2exp_models.audio2exp import Audio2Exp

I can only remember these changes I made. You may also need to dubug in some other places, but not much.

Well I tried but not that easy. Is it primarily based on old version and replacing(or adding) some files from new one? Or on new version and from new one?

[Mayy1994]

The sadtalker I used is v.0.0.1. The files in the folder "src" are copied from v.0.0.1, except for "utils".
The v.0.0.1 doesn't contain "utils", so "utils" is copied form v.0.0.2

[justin4ai]

The sadtalker I used is v.0.0.1. The files in the folder "src" are copied from v.0.0.1, except for "utils". The v.0.0.1 doesn't contain "utils", so "utils" is copied form v.0.0.2

Thank you so much for your help. After the weekends, I'm back at work and now came a lot more than before. I think all I need is properly mixing files, some from v0.0.1 and others from v0.0.2.

I am encountering these kinds of errors on and on, confused which file to use for the same functionality. So if you have time, would you run tree command under third/SadTalker then copy&paste all the results here? I know your time is precious so don't want to ask you one by one!

[justin4ai]

The sadtalker I used is v.0.0.1. The files in the folder "src" are copied from v.0.0.1, except for "utils". The v.0.0.1 doesn't contain "utils", so "utils" is copied form v.0.0.2

Actually I succeeded in inferencing, but the output sucks. What about yours?

[Mayy1994]

Glad to hear that.
I fed audio and a source image, and the generated talking head results are fine.

[justin4ai]

Glad to hear that.
I fed audio and a source image, and the generated talking head results are fine.

That's interesting. Do you mind sharing the code (the SadTalker folder you are using itself without checkpoints) by any means - ex. via Google drive or email? I want to make clean-code version of this repository such that many young learners perform their best in talking face generation task. Of course credit will also go to you too.

Anyways, appreciate your help even for a stranger. THANKS!

[mrlapuda]

I've change the code of inference.py :
`import argparse
import tqdm
import torch
import os
import utils.inference_util as inference_util
import utils.video_util as video_util
from utils.common import tensor2img
from configs.config import Config
from configs.path import PRETRAINED_MODELS_PATH
from models.styleheat.styleheat import StyleHEAT

def reenactment(generator, data):
"""
:param generator:
:param data: {
video_name: <class 'str'>
source_image: <class 'torch.Tensor'> (3, 256, 256)
source_semantics: <class 'torch.Tensor'> (73, 27)
target_image: <class 'list'> (B, 3, 256, 256)
target_semantics: <class 'list'> (B, 73, 27)
target_audio: <class 'list'> (B, 80, 16)
}
:return:
"""
bs = args.batch_size
source_image = data['source_align'].unsqueeze(0).cuda()
inv_data = inference_util.hfgi_inversion(generator, source_image, args=args, batch_size=bs)
source_image = source_image.repeat(bs, 1, 1, 1)

num_batch = len(data['target_semantics']) // bs + 1
gt_images, video_warp_images, audio_warp_images, fake_images = [], [], [], []
source_3dmm = data['source_semantics'].unsqueeze(-1).repeat(1, 1, 27)  # 1, 73, 27
for _i in range(num_batch):
    target_3dmm = data['target_semantics'][_i * bs:(_i + 1) * bs]
    if len(target_3dmm) == 0 or _i * bs > args.frame_limit:
        break
    
    if not torch.is_tensor(target_3dmm):
        target_3dmm = torch.stack(target_3dmm).cuda()
    else:
        target_3dmm = target_3dmm.cuda()

    if 'target_image' in data:
        target_images = data['target_image'][_i * bs:(_i + 1) * bs]
    else:
        target_images = None

    _len_3dmm = len(target_3dmm)
    if _len_3dmm < bs:
        # Last batch
        ix, wx, fx, inversion_condition = inv_data
        ix, wx, fx = ix[:_len_3dmm], wx[:_len_3dmm], fx[:_len_3dmm]
        if args.inversion_option == 'encode':
            inversion_condition = (inversion_condition[0][:_len_3dmm], inversion_condition[1][:_len_3dmm])
        inv_data = ix, wx, fx, inversion_condition

        source_3dmm = source_3dmm[:_len_3dmm]

    with torch.no_grad():
        if args.edit_expression_only:
            target_3dmm[:, 64:, :] = source_3dmm[:, 64:, :]
        output = generator.forward(source_image, target_3dmm, inv_data=inv_data, imsize=1024)

    # gt_images.append(target_images)
    fake_images.append(output['fake_image'].cpu().clamp_(-1, 1))
    # video_warp_images.append(output['video_warp_image'].cpu().clamp_(-1, 1))

fake_images = torch.cat(fake_images, 0)
# gt_images = torch.cat(gt_images, 0)
# video_warp_images = torch.cat(video_warp_images, 0)

video_util.write2video("{}/{}".format(args.output_dir, data['video_name']), fake_images)
print('Save video in {}/{}.mp4'.format(args.output_dir, data['video_name']))

def audio_reenactment(generator, data, audio_path):
# import TODO
import sys
import os
sys.path.append(os.path.abspath('./third_part/SadTalker/'))
import random
import numpy as np
from pydub import AudioSegment
from third_part.SadTalker.src.test_audio2coeff import Audio2Coeff
from third_part.SadTalker.src.generate_batch import get_data
from third_part.SadTalker.src.utils.preprocess import CropAndExtract
import scipy.io as scio
import warnings
warnings.filterwarnings("ignore")

current_root_path = os.path.abspath('./third_part/SadTalker/')
checkpoint_dir = './checkpoints/'
device = 'cuda'
temp_root = './docs/demo/output/temp/'
os.makedirs(temp_root, exist_ok=True)

wav2lip_checkpoint = os.path.join(current_root_path, checkpoint_dir, 'wav2lip.pth')

audio2pose_checkpoint = os.path.join(current_root_path, checkpoint_dir, 'auido2pose_00140-model.pth')
audio2pose_yaml_path = os.path.join(current_root_path, 'src', 'config', 'auido2pose.yaml')

audio2exp_checkpoint = os.path.join(current_root_path, checkpoint_dir, 'auido2exp_00300-model.pth')
audio2exp_yaml_path = os.path.join(current_root_path, 'src', 'config', 'auido2exp.yaml')

free_view_checkpoint = os.path.join(current_root_path, checkpoint_dir, 'facevid2vid_00189-model.pth.tar')
mapping_checkpoint = os.path.join(current_root_path, checkpoint_dir, 'mapping_00229-model.pth.tar')
facerender_yaml_path = os.path.join(current_root_path, 'src', 'config', 'facerender.yaml')
print(audio2pose_checkpoint)
sadtalker_path = {
'audio2pose_yaml_path': audio2pose_yaml_path,
'audio2exp_yaml_path': audio2exp_yaml_path,
'audio2pose_checkpoint': audio2pose_checkpoint,
'audio2exp_checkpoint': audio2exp_checkpoint,
'use_safetensor': False,  # 假设您不使用 safetensor，如果使用，请设置为 True
'checkpoint': None,  # 如果您使用 safetensor，请提供相应的路径

}

audio_to_coeff = Audio2Coeff(sadtalker_path, device)

first_coeff_path = os.path.join(temp_root, 'first_coeff.npy')
source_3dmm = data['source_semantics'][0, :-3].unsqueeze(0) # Reduce the crop params
crop_3dmm = data['source_semantics'][0, -3:].unsqueeze(0)
source_3dmm = source_3dmm.cpu().numpy()
# print(source_3dmm.shape)
# import pdb; pdb.set_trace()
scio.savemat(first_coeff_path, {'coeff_3dmm': source_3dmm})

'''
# path_of_lm_croper = os.path.join(current_root_path, checkpoint_dir, 'shape_predictor_68_face_landmarks.dat')
# path_of_net_recon_model = os.path.join(current_root_path, checkpoint_dir, 'epoch_20.pth')
# dir_of_BFM_fitting = os.path.join(current_root_path, checkpoint_dir, 'BFM_Fitting')
# preprocess_model = CropAndExtract(path_of_lm_croper, path_of_net_recon_model, dir_of_BFM_fitting, device)
# first_frame_dir = os.path.join(temp_root, 'first_frame_dir')
# os.makedirs(first_frame_dir, exist_ok=True)

# pic_path = os.path.join(first_frame_dir, 'first.jpg')
# tensor2img(data['source_align']).save(pic_path)

# first_coeff_path, crop_pic_path =  preprocess_model.generate(pic_path, first_frame_dir)
'''

audio_batch = get_data(first_coeff_path, audio_path, device,None,use_blink=False)
pose_style = random.randint(0, 45)
coeff_path = audio_to_coeff.generate(audio_batch, temp_root, pose_style)

audio_coeff = scio.loadmat(coeff_path)
audio_coeff = torch.from_numpy(audio_coeff['coeff_3dmm'])
audio_coeff = torch.cat([audio_coeff, crop_3dmm.repeat(audio_coeff.shape[0], 1)], dim=1)

# data['target_semantics'][0][:, 0][-3:]
# audio_coeff[0][-3:]
# import pdb; pdb.set_trace()
# print('Audio coeff shape: ', audio_coeff.shape)
semantic_radius = 13
def obtain_seq_index(index, num_frames):
    seq = list(range(index - semantic_radius, index + semantic_radius + 1))
    seq = [min(max(item, 0), num_frames - 1) for item in seq]
    return seq

def transform_semantic(semantic, frame_index):
    index = obtain_seq_index(frame_index, semantic.shape[0])
    coeff_3dmm = semantic[index, ...]
    return torch.Tensor(coeff_3dmm).permute(1, 0)

audio_coeff_list = []
for _i in range(len(audio_coeff)):
    audio_coeff_list.append(transform_semantic(audio_coeff, _i))
audio_coeff = torch.stack(audio_coeff_list)

data['target_semantics'] = audio_coeff.to(device)

reenactment(generator, data)
# import pdb; pdb.set_trace()
video_path = os.path.join(args.output_dir, data['image_name'] + '.mp4')

audio_path = audio_path
audio_name = os.path.splitext(os.path.split(audio_path)[-1])[0]
new_audio_path = os.path.join(args.output_dir, audio_name+'.wav')
start_time = 0
sound = AudioSegment.from_mp3(audio_path)

frames = len(data['target_semantics'])
end_time = start_time + frames*1/25*1000
word1 = sound.set_frame_rate(16000)
word = word1[start_time:end_time]
word.export(new_audio_path, format="wav")

av_path = os.path.join(args.output_dir, data['image_name'] + '_audio.mp4')
cmd = r'ffmpeg -y -i "%s" -i "%s" -vcodec copy "%s"' % (video_path, new_audio_path, av_path)
os.system(cmd)

def attribute_edit(generator, data):
assert args.attribute in ['young', 'old', 'beard', 'lip']
# Recommend factor
if args.attribute == 'young':
factor = -5.0
elif args.attribute == 'old':
factor = 5.0
elif args.attribute == 'beard':
factor = -20.0
elif args.attribute == 'lip':
factor = 20.0

bs = args.batch_size
source_image = data['source_align'].unsqueeze(0).cuda()
per_wx, per_ix, per_res = generator.generator.encode(source_image)

inv_data = inference_util.hfgi_inversion(generator, source_image, args=args, batch_size=bs)
source_image = source_image.repeat(bs, 1, 1, 1)

num_batch = len(data['target_image']) // bs + 1
gt_images, video_warp_images, audio_warp_images, fake_images = [], [], [], []
source_3dmm = data['source_semantics'].unsqueeze(-1).repeat(1, 1, 27)  # 1, 73, 27
for _i in range(num_batch):
    target_images = data['target_image'][_i * bs:(_i + 1) * bs]
    if len(target_images) == 0 or _i * bs > args.frame_limit:
        break
    target_3dmm = data['target_semantics'][_i * bs:(_i + 1) * bs]
    target_3dmm = torch.stack(target_3dmm).cuda()

    _len_3dmm = len(target_3dmm)
    if _len_3dmm < bs:
        # Last batch
        ix, wx, fx, inversion_condition = inv_data
        ix, wx, fx = ix[:_len_3dmm], wx[:_len_3dmm], fx[:_len_3dmm]
        if args.inversion_option == 'encode':
            inversion_condition = (inversion_condition[0][:_len_3dmm], inversion_condition[1][:_len_3dmm])
        inv_data = ix, wx, fx, inversion_condition
        source_3dmm = source_3dmm[:_len_3dmm]

    with torch.no_grad():
        if args.edit_expression_only:
            target_3dmm[:, 64:, :] = source_3dmm[:, 64:, :]
        output = generator.forward(source_image, target_3dmm, inv_data=inv_data, imsize=1024)

    ix_edit, wx_edit, fx_edit, inversion_condition = generator. \
        generator.edit(x=None, factor=factor / num_batch * (_i + 1), choice=args.attribute, wx=per_wx, res=per_res)
    inv_data = [
        ix_edit.expand(bs, 3, 256, 256),
        wx_edit.expand(bs, 18, 512),
        fx_edit.expand(bs, 512, 64, 64),
        (inversion_condition[0].expand(bs, 512, 64, 64),
         inversion_condition[1].expand(bs, 512, 64, 64))
    ]

    gt_images.append(target_images)
    fake_images.append(output['fake_image'].cpu().clamp_(-1, 1))
    video_warp_images.append(output['video_warp_image'].cpu().clamp_(-1, 1))

fake_images = torch.cat(fake_images, 0)
gt_images = torch.cat(gt_images, 0)
video_warp_images = torch.cat(video_warp_images, 0)

video_util.write2video("{}/{}".format(args.output_dir, data['video_name']) + '_attribute_edit', fake_images)

def intuitive_edit(generator, data):
source_image = data['source_align'].unsqueeze(0).cuda()
inv_data = inference_util.hfgi_inversion(generator, source_image, args=args, batch_size=1)

control_dict, sort_rot_control, sort_exp_control = inference_util.get_intuitive_control()
step = 10
output_images = []
# rotation control
current = control_dict['rotation_center']
data['source_semantics'] = data['source_semantics'].unsqueeze(-1).repeat(1, 1, 27)
for control in sort_rot_control:
    rotation = None
    for i in range(step):
        rotation = (control_dict[control] - current) * i / (step - 1) + current
        data['source_semantics'][:, 64:70, :] = rotation[None, :, None]
        with torch.no_grad():
            output = generator.forward(source_image, data['source_semantics'].cuda(), inv_data=inv_data, imsize=1024)
        output_images.append(output['fake_image'].cpu().clamp_(-1, 1))
    current = rotation

# expression control
current = data['source_semantics'][0, :64, 0]
for control in sort_exp_control:
    expression = None
    for i in range(step):
        expression = (control_dict[control] - current) * i / (step - 1) + current
        data['source_semantics'][:, :64, :] = expression[None, :, None]
        with torch.no_grad():
            output = generator.forward(source_image, data['source_semantics'].cuda(), inv_data=inv_data, imsize=1024)
        output_images.append(output['fake_image'].cpu().clamp_(-1, 1))
    current = expression
output_images = torch.cat(output_images, 0)

video_util.write2video("{}/{}".format(args.output_dir, data['image_name']) + '_intuitive_edit', output_images)

def parse_args():
parser = argparse.ArgumentParser(description='Inferencing')
parser.add_argument('--config', default='./configs/inference.yaml')
parser.add_argument('--name', default='test')
parser.add_argument('--from_dataset', action='store_true')
parser.add_argument('--cross_id', action='store_true')
parser.add_argument('--image_source', type=str, default=None, help='Single path or directory')
parser.add_argument('--video_source', type=str, default=None, help='Single path or directory')
parser.add_argument('--output_dir', default='./')
parser.add_argument('--inversion_option', type=str, default='encode', help='load, optimize, encode')
parser.add_argument('--frame_limit', type=int, default=1000)
parser.add_argument('--batch_size', type=int, default=10)
parser.add_argument('--if_extract', action='store_true')
parser.add_argument('--if_align', action='store_true')
# Audio
parser.add_argument('--enable_audio', action='store_true')
parser.add_argument('--audio_path', type=str, default=None)
# Editing
parser.add_argument('--attribute_edit', action='store_true')
parser.add_argument('--attribute', type=str, default=None)
parser.add_argument('--intuitive_edit', action='store_true')
parser.add_argument('--edit_expression_only', action='store_true')
args = parser.parse_args()
return args

def main():
opt = Config(args.config)
opt.model.enable_audio = args.enable_audio
generator = StyleHEAT(opt.model, PRETRAINED_MODELS_PATH).cuda()
dataset = inference_util.build_inference_dataset(args, opt)

for _ in tqdm.tqdm(range(len(dataset))):
    data = dataset.load_next_video()
    if args.intuitive_edit:
        intuitive_edit(generator, data)
    elif args.attribute_edit:
        attribute_edit(generator, data)
    elif args.audio_path is not None:
        audio_reenactment(generator, data, args.audio_path)
    else:
        reenactment(generator, data)

if name == 'main':
args = parse_args()
if args.cross_id:
print('Cross-id testing')
else:
print('Same-id testing')
if not os.path.exists(args.output_dir):
os.makedirs(args.output_dir)
main()

`
and your should download model from https://huggingface.co/vinthony/SadTalker/ if neeed.because some model may needed during execing

[mrlapuda]

key point is
` ``sadtalker_path = {
'audio2pose_yaml_path': audio2pose_yaml_path,
'audio2exp_yaml_path': audio2exp_yaml_path,
'audio2pose_checkpoint': audio2pose_checkpoint,
'audio2exp_checkpoint': audio2exp_checkpoint,
'use_safetensor': False, # 假设您不使用 safetensor，如果使用，请设置为 True
'checkpoint': None, # 如果您使用 safetensor，请提供相应的路径
}

audio_to_coeff = Audio2Coeff(sadtalker_path, device)

the way call function  Audio2Coeff has been changed