Spaces:
Sleeping
Sleeping
add total variation loss + tuning changes
Browse files
app.py
CHANGED
@@ -11,14 +11,19 @@ model = ViTForImageClassification.from_pretrained('google/vit-large-patch32-384'
|
|
11 |
model.to(device)
|
12 |
model.eval()
|
13 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
14 |
def process_image(input_image, learning_rate, iterations, n_targets, seed):
|
15 |
if input_image is None:
|
16 |
return None
|
17 |
-
|
18 |
-
def get_encoder_activations(x):
|
19 |
-
encoder_output = model.vit(x)
|
20 |
-
final_activations = encoder_output.last_hidden_state[:,0,:]
|
21 |
-
return final_activations
|
22 |
|
23 |
image = input_image.convert('RGB')
|
24 |
pixel_values = feature_extractor(images=image, return_tensors="pt").pixel_values
|
@@ -36,8 +41,11 @@ def process_image(input_image, learning_rate, iterations, n_targets, seed):
|
|
36 |
|
37 |
final_activations = get_encoder_activations(pixel_values)
|
38 |
logits = model.classifier(final_activations[0])
|
39 |
-
|
40 |
-
|
|
|
|
|
|
|
41 |
|
42 |
with torch.no_grad():
|
43 |
pixel_values.data += learning_rate * pixel_values.grad.data
|
@@ -52,9 +60,10 @@ iface = gr.Interface(
|
|
52 |
fn=process_image,
|
53 |
inputs=[
|
54 |
gr.Image(type="pil"),
|
55 |
-
gr.Number(value=
|
56 |
-
gr.Number(value=
|
57 |
-
gr.Number(value=
|
|
|
58 |
gr.Number(value=50, minimum=1, maximum=1000, label="Number of Random Target Class Activations to Maximise"),
|
59 |
],
|
60 |
outputs=[gr.Image(type="numpy", label="ViT-Dreamed Image")]
|
|
|
11 |
model.to(device)
|
12 |
model.eval()
|
13 |
|
14 |
+
def get_encoder_activations(x):
|
15 |
+
encoder_output = model.vit(x)
|
16 |
+
final_activations = encoder_output.last_hidden_state[:,0,:]
|
17 |
+
return final_activations
|
18 |
+
|
19 |
+
def total_variation_loss(img):
|
20 |
+
pixel_dif1 = img[:, :, 1:, :] - img[:, :, :-1, :]
|
21 |
+
pixel_dif2 = img[:, :, :, 1:] - img[:, :, :, :-1]
|
22 |
+
return (torch.sum(torch.abs(pixel_dif1)) + torch.sum(torch.abs(pixel_dif2)))
|
23 |
+
|
24 |
def process_image(input_image, learning_rate, iterations, n_targets, seed):
|
25 |
if input_image is None:
|
26 |
return None
|
|
|
|
|
|
|
|
|
|
|
27 |
|
28 |
image = input_image.convert('RGB')
|
29 |
pixel_values = feature_extractor(images=image, return_tensors="pt").pixel_values
|
|
|
41 |
|
42 |
final_activations = get_encoder_activations(pixel_values)
|
43 |
logits = model.classifier(final_activations[0])
|
44 |
+
|
45 |
+
original_loss = -logits[random_indices].sum()
|
46 |
+
tv_loss = total_variation_loss(pixel_values)
|
47 |
+
total_loss = original_loss + 0.00625 * tv_loss
|
48 |
+
total_loss.backward()
|
49 |
|
50 |
with torch.no_grad():
|
51 |
pixel_values.data += learning_rate * pixel_values.grad.data
|
|
|
60 |
fn=process_image,
|
61 |
inputs=[
|
62 |
gr.Image(type="pil"),
|
63 |
+
gr.Number(value=10.0, minimum=0, label="Learning Rate"),
|
64 |
+
gr.Number(value=0.00625, label="Total Variation Loss"),
|
65 |
+
gr.Number(value=1, minimum=1, label="Iterations"),
|
66 |
+
gr.Number(value=420, minimum=0, label="Seed"),
|
67 |
gr.Number(value=50, minimum=1, maximum=1000, label="Number of Random Target Class Activations to Maximise"),
|
68 |
],
|
69 |
outputs=[gr.Image(type="numpy", label="ViT-Dreamed Image")]
|