Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -12,15 +12,14 @@ from Scripts.model import create_cam, create_model
|
|
12 |
from Scripts.preprocess import crop_face, extract_face, extract_frames
|
13 |
from Scripts.ca_generator import get_augs
|
14 |
|
15 |
-
|
16 |
-
|
17 |
|
|
|
18 |
|
19 |
-
device = torch.device('cpu')
|
20 |
|
21 |
sbcl = create_model("Weights/weights.tar")
|
22 |
|
23 |
-
|
24 |
face_detector = get_model("resnet50_2020-07-20", max_size=1024, device=device)
|
25 |
face_detector.eval()
|
26 |
|
@@ -37,9 +36,8 @@ dlib_face_detector = dlib.get_frontal_face_detector()
|
|
37 |
dlib_face_predictor = dlib.shape_predictor(
|
38 |
'Weights/shape_predictor_81_face_landmarks.dat')
|
39 |
|
40 |
-
|
41 |
def predict_image(inp):
|
42 |
-
|
43 |
face_list = extract_face(inp, face_detector)
|
44 |
|
45 |
if len(face_list) == 0:
|
@@ -56,9 +54,8 @@ def predict_image(inp):
|
|
56 |
|
57 |
return confidences, cam_image
|
58 |
|
59 |
-
|
60 |
def predict_video(inp):
|
61 |
-
|
62 |
face_list, idx_list = extract_frames(inp, 10, face_detector)
|
63 |
|
64 |
with torch.no_grad():
|
@@ -84,8 +81,6 @@ def predict_video(inp):
|
|
84 |
|
85 |
return {'Real': 1-pred, 'Fake': pred}, cam_image
|
86 |
|
87 |
-
|
88 |
-
|
89 |
with gr.Blocks(title="Deepfake Detection CL", theme='upsatwal/mlsc_tiet', css="""
|
90 |
@import url('https://fonts.googleapis.com/css?family=Source+Code+Pro:200');
|
91 |
#custom_header {
|
@@ -186,4 +181,4 @@ with gr.Blocks(title="Deepfake Detection CL", theme='upsatwal/mlsc_tiet', css=""
|
|
186 |
btn_video.click(predict_video, inputs=input_video, outputs=[label_probs_video, output_image_video], api_name="/predict_video")
|
187 |
|
188 |
if __name__ == "__main__":
|
189 |
-
demo.launch()
|
|
|
12 |
from Scripts.preprocess import crop_face, extract_face, extract_frames
|
13 |
from Scripts.ca_generator import get_augs
|
14 |
|
15 |
+
import spaces
|
|
|
16 |
|
17 |
+
warnings.filterwarnings('ignore')
|
18 |
|
19 |
+
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
|
20 |
|
21 |
sbcl = create_model("Weights/weights.tar")
|
22 |
|
|
|
23 |
face_detector = get_model("resnet50_2020-07-20", max_size=1024, device=device)
|
24 |
face_detector.eval()
|
25 |
|
|
|
36 |
dlib_face_predictor = dlib.shape_predictor(
|
37 |
'Weights/shape_predictor_81_face_landmarks.dat')
|
38 |
|
39 |
+
@spaces.GPU
|
40 |
def predict_image(inp):
|
|
|
41 |
face_list = extract_face(inp, face_detector)
|
42 |
|
43 |
if len(face_list) == 0:
|
|
|
54 |
|
55 |
return confidences, cam_image
|
56 |
|
57 |
+
@spaces.GPU
|
58 |
def predict_video(inp):
|
|
|
59 |
face_list, idx_list = extract_frames(inp, 10, face_detector)
|
60 |
|
61 |
with torch.no_grad():
|
|
|
81 |
|
82 |
return {'Real': 1-pred, 'Fake': pred}, cam_image
|
83 |
|
|
|
|
|
84 |
with gr.Blocks(title="Deepfake Detection CL", theme='upsatwal/mlsc_tiet', css="""
|
85 |
@import url('https://fonts.googleapis.com/css?family=Source+Code+Pro:200');
|
86 |
#custom_header {
|
|
|
181 |
btn_video.click(predict_video, inputs=input_video, outputs=[label_probs_video, output_image_video], api_name="/predict_video")
|
182 |
|
183 |
if __name__ == "__main__":
|
184 |
+
demo.launch()
|