-
Notifications
You must be signed in to change notification settings - Fork 6
/
Copy pathinference.py
141 lines (129 loc) · 7.56 KB
/
inference.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
import os
import cv2
import glob,copy
import shutil
from tqdm import tqdm
import torch
import pickle
import numpy as np
from cuda_malloc import cuda_malloc_supported
from typing import Any
import folder_paths
from .musetalk.utils.face_parsing import FaceParsing
from mmpose.apis import init_model
from .musetalk.utils.blending import get_image
from .musetalk.utils.utils import load_all_model,get_file_type,get_video_fps,datagen
from .musetalk.utils.preprocessing import get_landmark_and_bbox,read_imgs,coord_placeholder
parent_directory = os.path.dirname(os.path.abspath(__file__))
input_path = folder_paths.get_input_directory()
out_path = folder_paths.get_output_directory()
class MuseTalk_INFER:
def __init__(self,bbox_shift=0,fps=25,
batch_size=8,batch_size_fa=2,
use_saved_coord=False) -> None:
self.fps = fps
self.bbox_shift = bbox_shift
self.batch_size = batch_size
self.batch_size_fa = batch_size_fa
self.use_saved_coord = use_saved_coord
self.device = torch.device("cuda" if cuda_malloc_supported() else "cpu")
config_file = os.path.join(parent_directory,"musetalk","utils","dwpose","rtmpose-l_8xb32-270e_coco-ubody-wholebody-384x288.py")
checkpoint_file = os.path.join(parent_directory,'models','dwpose','dw-ll_ucoco_384.pth')
resnet_path = os.path.join(parent_directory,'models','face-parse-bisent','resnet18-5c106cde.pth')
face_model_pth = os.path.join(parent_directory,"models','face-parse-bisent','79999_iter.pth")
self.fp_model = FaceParsing(resnet_path,face_model_pth)
self.dwpose_model = init_model(config_file, checkpoint_file, device=self.device)
self.audio_processor,self.vae,self.unet,self.pe = load_all_model(os.path.join(parent_directory,"models"))
self.timesteps = torch.tensor([0], device=self.device)
def __call__(self, video_path,audio_path,*args: Any, **kwds: Any) -> Any:
input_basename = os.path.basename(video_path).split('.')[0]
audio_basename = os.path.basename(audio_path).split('.')[0]
output_basename = f"{input_basename}_{audio_basename}"
result_img_save_path = os.path.join(out_path,"musetalk_result", output_basename) # related to video & audio inputs
os.makedirs(result_img_save_path, exist_ok=True)
crop_coord_save_path = os.path.join(result_img_save_path, input_basename+".pkl") # only related to video input
output_vid_name = os.path.join(out_path, output_basename+".mp4")
############################################## extract frames from source video ##############################################
if get_file_type(video_path)=="video":
save_dir_full = os.path.join(out_path,"musetalk_result",input_basename)
os.makedirs(save_dir_full,exist_ok = True)
png_path = os.path.join(save_dir_full,"%08d.png")
cmd = f"ffmpeg -v fatal -i {video_path} -start_number 0 {png_path}"
os.system(cmd)
input_img_list = sorted(glob.glob(os.path.join(save_dir_full, '*.[jpJP][pnPN]*[gG]')))
fps = get_video_fps(video_path)
else: # input img folder
input_img_list = glob.glob(os.path.join(video_path, '*.[jpJP][pnPN]*[gG]'))
input_img_list = sorted(input_img_list, key=lambda x: int(os.path.splitext(os.path.basename(x))[0]))
fps = self.fps
#print(input_img_list)
############################################## extract audio feature ##############################################
whisper_feature = self.audio_processor.audio2feat(audio_path)
whisper_chunks = self.audio_processor.feature2chunks(feature_array=whisper_feature,fps=fps)
############################################## preprocess input image ##############################################
if os.path.exists(crop_coord_save_path) and self.use_saved_coord:
print("using extracted coordinates")
with open(crop_coord_save_path,'rb') as f:
coord_list = pickle.load(f)
frame_list = read_imgs(input_img_list)
else:
print("extracting landmarks...time consuming")
coord_list, frame_list = get_landmark_and_bbox(self.dwpose_model,input_img_list, self.batch_size_fa,self.bbox_shift)
with open(crop_coord_save_path, 'wb') as f:
pickle.dump(coord_list, f)
i = 0
input_latent_list = []
for bbox, frame in zip(coord_list, frame_list):
if bbox == coord_placeholder:
continue
x1, y1, x2, y2 = bbox
crop_frame = frame[y1:y2, x1:x2]
crop_frame = cv2.resize(crop_frame,(256,256),interpolation = cv2.INTER_LANCZOS4)
latents = self.vae.get_latents_for_unet(crop_frame)
input_latent_list.append(latents)
# to smooth the first and the last frame
frame_list_cycle = frame_list + frame_list[::-1]
coord_list_cycle = coord_list + coord_list[::-1]
input_latent_list_cycle = input_latent_list + input_latent_list[::-1]
############################################## inference batch by batch ##############################################
print("start inference")
video_num = len(whisper_chunks)
batch_size = self.batch_size
gen = datagen(whisper_chunks,input_latent_list_cycle,batch_size)
res_frame_list = []
for i, (whisper_batch,latent_batch) in enumerate(tqdm(gen,total=int(np.ceil(float(video_num)/batch_size)))):
tensor_list = [torch.FloatTensor(arr) for arr in whisper_batch]
audio_feature_batch = torch.stack(tensor_list).to(self.unet.device) # torch, B, 5*N,384
audio_feature_batch = self.pe(audio_feature_batch)
pred_latents = self.unet.model(latent_batch, self.timesteps, encoder_hidden_states=audio_feature_batch).sample
recon = self.vae.decode_latents(pred_latents)
for res_frame in recon:
res_frame_list.append(res_frame)
############################################## pad to full image ##############################################
print("pad talking image to original video")
for i, res_frame in enumerate(tqdm(res_frame_list)):
bbox = coord_list_cycle[i%(len(coord_list_cycle))]
ori_frame = copy.deepcopy(frame_list_cycle[i%(len(frame_list_cycle))])
x1, y1, x2, y2 = bbox
try:
res_frame = cv2.resize(res_frame.astype(np.uint8),(x2-x1,y2-y1))
except:
# print(bbox)
continue
combine_frame = get_image(self.fp_model, ori_frame,res_frame,bbox)
cv2.imwrite(os.path.join(result_img_save_path,f"{str(i).zfill(8)}.png"),combine_frame)
res_tmp_path = os.path.join(result_img_save_path,"%08d.png")
cmd_img2video = f"ffmpeg -y -v fatal -r {fps} -f image2 -i {res_tmp_path} -vcodec libx264 -vf format=rgb24,scale=out_color_matrix=bt709,format=yuv420p -crf 18 {output_vid_name}"
print(cmd_img2video)
os.system(cmd_img2video)
'''
cmd_combine_audio = f"ffmpeg -y -v fatal -i {audio_path} -i temp.mp4 {output_vid_name}"
print(cmd_combine_audio)
os.system(cmd_combine_audio)
'''
del self.fp_model,self.dwpose_model,self.audio_processor,self.vae,self.unet,self.pe
torch.cuda.empty_cache()
# os.remove("temp.mp4")
shutil.rmtree(result_img_save_path)
print(f"result is save to {output_vid_name}")
return output_vid_name