mahmudunnabi
commited on
Commit
•
1db4907
1
Parent(s):
7f1929a
Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,107 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Import the dependencies
|
2 |
+
import gradio as gr
|
3 |
+
from PIL import Image
|
4 |
+
import torch
|
5 |
+
from transformers import SamModel, SamProcessor
|
6 |
+
import numpy as np
|
7 |
+
import matplotlib.pyplot as plt
|
8 |
+
|
9 |
+
|
10 |
+
# Load the SAM model and processor
|
11 |
+
model = SamModel.from_pretrained("Zigeng/SlimSAM-uniform-77")
|
12 |
+
processor = SamProcessor.from_pretrained("Zigeng/SlimSAM-uniform-77")
|
13 |
+
|
14 |
+
|
15 |
+
# Global variable to store input points
|
16 |
+
input_points = []
|
17 |
+
|
18 |
+
# Helper functions
|
19 |
+
def show_mask(mask, ax, random_color=False):
|
20 |
+
if random_color:
|
21 |
+
color = np.concatenate([np.random.random(3),
|
22 |
+
np.array([0.6])],
|
23 |
+
axis=0)
|
24 |
+
else:
|
25 |
+
color = np.array([30/255, 144/255, 255/255, 0.6])
|
26 |
+
h, w = mask.shape[-2:]
|
27 |
+
mask_image = mask.reshape(h, w, 1) * color.reshape(1, 1, -1)
|
28 |
+
ax.imshow(mask_image)
|
29 |
+
# Function to get pixel coordinates
|
30 |
+
def get_pixel_coordinates(image, evt: gr.SelectData):
|
31 |
+
global input_points
|
32 |
+
x, y = evt.index[0], evt.index[1]
|
33 |
+
input_points = [[[x, y]]]
|
34 |
+
return perform_prediction(image)
|
35 |
+
|
36 |
+
# Function to perform SAM model prediction
|
37 |
+
def perform_prediction(image):
|
38 |
+
global input_points
|
39 |
+
# Preprocess the image
|
40 |
+
inputs = processor(images=image, input_points=input_points, return_tensors="pt")
|
41 |
+
# Perform prediction
|
42 |
+
with torch.no_grad():
|
43 |
+
outputs = model(**inputs)
|
44 |
+
iou = outputs.iou_scores
|
45 |
+
max_iou_index = torch.argmax(iou)
|
46 |
+
|
47 |
+
# Post-process the masks
|
48 |
+
predicted_masks = processor.image_processor.post_process_masks(
|
49 |
+
outputs.pred_masks,
|
50 |
+
inputs['original_sizes'],
|
51 |
+
inputs['reshaped_input_sizes']
|
52 |
+
)
|
53 |
+
predicted_mask = predicted_masks[0]
|
54 |
+
|
55 |
+
# Display the mask on the image
|
56 |
+
mask_image = show_mask_on_image(image, predicted_mask[:,max_iou_index], return_image=True)
|
57 |
+
return mask_image
|
58 |
+
|
59 |
+
# Function to overlay mask on the image
|
60 |
+
def show_mask_on_image(raw_image, mask, return_image=False):
|
61 |
+
if not isinstance(mask, torch.Tensor):
|
62 |
+
mask = torch.Tensor(mask)
|
63 |
+
|
64 |
+
if len(mask.shape) == 4:
|
65 |
+
mask = mask.squeeze()
|
66 |
+
|
67 |
+
fig, axes = plt.subplots(1, 1, figsize=(15, 15))
|
68 |
+
|
69 |
+
mask = mask.cpu().detach()
|
70 |
+
axes.imshow(np.array(raw_image))
|
71 |
+
show_mask(mask, axes)
|
72 |
+
axes.axis("off")
|
73 |
+
plt.show()
|
74 |
+
|
75 |
+
if return_image:
|
76 |
+
fig = plt.gcf()
|
77 |
+
fig.canvas.draw()
|
78 |
+
# Convert plot to image
|
79 |
+
img = np.frombuffer(fig.canvas.tostring_rgb(), dtype=np.uint8)
|
80 |
+
img = img.reshape(fig.canvas.get_width_height()[::-1] + (3,))
|
81 |
+
img = Image.fromarray(img)
|
82 |
+
plt.close(fig)
|
83 |
+
return img
|
84 |
+
|
85 |
+
|
86 |
+
|
87 |
+
# Create the Gradio interface
|
88 |
+
with gr.Blocks() as demo:
|
89 |
+
gr.Markdown(
|
90 |
+
"""
|
91 |
+
<div style='text-align: center; font-family: "Times New Roman";'>
|
92 |
+
<h1 style='color: #FF6347;'>One Click Image Segmentation App</h1>
|
93 |
+
<h3 style='color: #4682B4;'>Model: SlimSAM-uniform-77</h3>
|
94 |
+
<h3 style='color: #32CD32;'>Made By: Md. Mahmudun Nabi</h3>
|
95 |
+
</div>
|
96 |
+
"""
|
97 |
+
)
|
98 |
+
with gr.Row():
|
99 |
+
|
100 |
+
img = gr.Image(type="pil", label="Input Image",height=400, width=600)
|
101 |
+
output_image = gr.Image(label="Masked Image")
|
102 |
+
|
103 |
+
img.select(get_pixel_coordinates, inputs=[img], outputs=[output_image])
|
104 |
+
|
105 |
+
|
106 |
+
if __name__ == "__main__":
|
107 |
+
demo.launch(share=False)
|