Spaces:
Starting
Starting
Felix Konrad
commited on
Commit
·
2ec5753
1
Parent(s):
57c8491
Added proper Cosine-Similarity Computation + Visualization
Browse files
app.py
CHANGED
@@ -1,4 +1,5 @@
|
|
1 |
import matplotlib.pyplot as plt
|
|
|
2 |
import numpy as np
|
3 |
import gradio as gr
|
4 |
from transformers import AutoModel, AutoImageProcessor
|
@@ -7,29 +8,68 @@ import torch
|
|
7 |
|
8 |
# Global state to store loaded model + processor
|
9 |
state = {
|
|
|
10 |
"model": None,
|
11 |
"processor": None,
|
12 |
"repo_id": None,
|
13 |
}
|
14 |
|
15 |
|
16 |
-
def
|
17 |
"""
|
18 |
-
|
19 |
-
Returns a PIL image that can be displayed in Gradio
|
20 |
"""
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
25 |
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
img = img.reshape(fig.canvas.get_width_height()[::-1] + (3,))
|
30 |
|
31 |
-
|
32 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
33 |
|
34 |
|
35 |
def load_model(repo_id: str, revision: str = None):
|
@@ -44,6 +84,8 @@ def load_model(repo_id: str, revision: str = None):
|
|
44 |
model.to("cuda")
|
45 |
else:
|
46 |
model.to("cpu")
|
|
|
|
|
47 |
# Store in global state
|
48 |
state["model"] = model
|
49 |
state["processor"] = processor
|
@@ -58,10 +100,21 @@ def display_image(image: Image):
|
|
58 |
"""
|
59 |
return image
|
60 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
61 |
# Build the Gradio interface
|
62 |
with gr.Blocks() as demo:
|
63 |
gr.Markdown("# Dynamic ViT Loader Template")
|
64 |
|
|
|
|
|
|
|
65 |
with gr.Row():
|
66 |
repo_input = gr.Textbox(label="Hugging Face model repo ID", placeholder="e.g. google/vit-base-patch16-224")
|
67 |
revision_input = gr.Textbox(label="Revision (optional)", placeholder="branch, tag, or commit hash")
|
@@ -72,13 +125,13 @@ with gr.Blocks() as demo:
|
|
72 |
image_output = gr.Image(label="Displayed Image")
|
73 |
|
74 |
# cos-sim visualization:
|
75 |
-
|
76 |
-
sim_array = np.random.normal((128, 128))
|
77 |
-
heatmap_img = plot_similarity_heatmap(sim_array)
|
78 |
-
gr.Image(value=heatmap_img, label="Cosine Similarity Heatmap")
|
79 |
|
80 |
# Button clicks / image upload handlers
|
81 |
load_btn.click(fn=load_model, inputs=[repo_input, revision_input], outputs=load_status)
|
82 |
image_input.change(fn=display_image, inputs=image_input, outputs=image_output)
|
83 |
|
|
|
|
|
|
|
84 |
demo.launch()
|
|
|
1 |
import matplotlib.pyplot as plt
|
2 |
+
import matplotlib.cm as cm
|
3 |
import numpy as np
|
4 |
import gradio as gr
|
5 |
from transformers import AutoModel, AutoImageProcessor
|
|
|
8 |
|
9 |
# Global state to store loaded model + processor
|
10 |
state = {
|
11 |
+
"model_type": None,
|
12 |
"model": None,
|
13 |
"processor": None,
|
14 |
"repo_id": None,
|
15 |
}
|
16 |
|
17 |
|
18 |
+
def similarity_heatmap(image):
|
19 |
"""
|
20 |
+
...
|
|
|
21 |
"""
|
22 |
+
model, processor = state["model"], state["processor"]
|
23 |
+
|
24 |
+
inputs = processor(images=image, return_tensors="pt")
|
25 |
+
pixel_values = inputs["pixel_values"].to(model.device) # shape: (1, 3, H, W)
|
26 |
+
|
27 |
+
# get ViT patch size (from model config)
|
28 |
+
patch_size = model.config.patch_size # usually 16
|
29 |
+
|
30 |
+
# Compute patch grid (needed for resizing later)
|
31 |
+
H_patch = pixel_values.shape[2] // patch_size
|
32 |
+
W_patch = pixel_values.shape[3] // patch_size
|
33 |
+
|
34 |
+
with torch.no_grad():
|
35 |
+
outputs = model(pixel_values) # last_hidden_state: (1, seq_len, hidden_dim)
|
36 |
+
last_hidden_state = outputs.last_hidden_state
|
37 |
+
cls_token = last_hidden_state[:, 0, :] # shape: (1, hidden_dim)
|
38 |
+
patch_tokens = last_hidden_state[:, 1:, :] # shape: (1, num_patches, hidden_dim)
|
39 |
+
|
40 |
+
cls_norm = cls_token / cls_token.norm(dim=-1, keepdim=True)
|
41 |
+
patch_norm = patch_tokens / patch_tokens.norm(dim=-1, keepdim=True)
|
42 |
+
|
43 |
+
cos_sim = torch.einsum("bd,bpd->bp", cls_norm, patch_norm) # shape: (1, num_patches)
|
44 |
+
cos_sim = cos_sim.reshape((H_patch, W_patch))
|
45 |
+
return np.array(cos_sim)
|
46 |
+
|
47 |
+
def overlay_cosine_grid_on_image(cos_grid: np.ndarray, image: Image.Image, alpha=0.5, colormap="viridis"):
|
48 |
+
"""
|
49 |
+
cos_grid: (H_patch, W_patch) numpy array of cosine similarities
|
50 |
+
image: PIL.Image
|
51 |
+
alpha: blending factor
|
52 |
+
colormap: matplotlib colormap name
|
53 |
+
"""
|
54 |
+
# Normalize cosine values to [0, 1] for colormap
|
55 |
+
norm_grid = (cos_grid - cos_grid.min()) / (cos_grid.max() - cos_grid.min() + 1e-8)
|
56 |
+
|
57 |
+
# Apply colormap
|
58 |
+
cmap = cm.get_cmap(colormap)
|
59 |
+
heatmap_rgba = cmap(norm_grid) # shape: (H_patch, W_patch, 4)
|
60 |
|
61 |
+
# Convert to RGB 0-255
|
62 |
+
heatmap_rgb = (heatmap_rgba[:, :, :3] * 255).astype(np.uint8)
|
63 |
+
heatmap_img = Image.fromarray(heatmap_rgb)
|
|
|
64 |
|
65 |
+
# Resize heatmap to match original image size
|
66 |
+
heatmap_resized = heatmap_img.resize(image.size, resample=Image.BILINEAR)
|
67 |
+
|
68 |
+
# Blend with original image
|
69 |
+
blended = Image.blend(image.convert("RGBA"), heatmap_resized.convert("RGBA"), alpha=alpha)
|
70 |
+
|
71 |
+
return blended
|
72 |
+
|
73 |
|
74 |
|
75 |
def load_model(repo_id: str, revision: str = None):
|
|
|
84 |
model.to("cuda")
|
85 |
else:
|
86 |
model.to("cpu")
|
87 |
+
|
88 |
+
model.eval()
|
89 |
# Store in global state
|
90 |
state["model"] = model
|
91 |
state["processor"] = processor
|
|
|
100 |
"""
|
101 |
return image
|
102 |
|
103 |
+
def visualize_cosine_heatmap(image: Image):
|
104 |
+
if state["model"] is None:
|
105 |
+
return None # or placeholder image
|
106 |
+
|
107 |
+
cos_grid = similarity_heatmap(image)
|
108 |
+
blended = overlay_cosine_grid_on_image(cos_grid, image)
|
109 |
+
return blended
|
110 |
+
|
111 |
# Build the Gradio interface
|
112 |
with gr.Blocks() as demo:
|
113 |
gr.Markdown("# Dynamic ViT Loader Template")
|
114 |
|
115 |
+
# TODO: Add drop-down menu (or something else) for user to allow choosing model type (e.g. DINOv2, Google ViT-Base etc.)
|
116 |
+
# ...
|
117 |
+
|
118 |
with gr.Row():
|
119 |
repo_input = gr.Textbox(label="Hugging Face model repo ID", placeholder="e.g. google/vit-base-patch16-224")
|
120 |
revision_input = gr.Textbox(label="Revision (optional)", placeholder="branch, tag, or commit hash")
|
|
|
125 |
image_output = gr.Image(label="Displayed Image")
|
126 |
|
127 |
# cos-sim visualization:
|
128 |
+
heatmap_output = gr.Image(label="Cosine Similarity Heatmap")
|
|
|
|
|
|
|
129 |
|
130 |
# Button clicks / image upload handlers
|
131 |
load_btn.click(fn=load_model, inputs=[repo_input, revision_input], outputs=load_status)
|
132 |
image_input.change(fn=display_image, inputs=image_input, outputs=image_output)
|
133 |
|
134 |
+
compute_btn = gr.Button("Compute Heatmap")
|
135 |
+
compute_btn.click(fn=visualize_cosine_heatmap, inputs=image_input, outputs=heatmap_output)
|
136 |
+
|
137 |
demo.launch()
|