Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -5,6 +5,7 @@ import torch
|
|
5 |
|
6 |
import matplotlib.pyplot as plt
|
7 |
import numpy as np
|
|
|
8 |
|
9 |
import gradio as gr
|
10 |
|
@@ -14,15 +15,20 @@ processor = CLIPSegProcessor.from_pretrained("CIDAS/clipseg-rd64-refined")
|
|
14 |
model = CLIPSegForImageSegmentation.from_pretrained("CIDAS/clipseg-rd64-refined")
|
15 |
|
16 |
def visualize_segmentation(image, prompts, preds):
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
22 |
|
23 |
|
24 |
def segment(img, clases):
|
25 |
-
print(img)
|
26 |
image = Image.fromarray(img, 'RGB')
|
27 |
prompts = clases.split(',')
|
28 |
|
@@ -31,7 +37,7 @@ def segment(img, clases):
|
|
31 |
outputs = model(**inputs)
|
32 |
preds = outputs.logits.unsqueeze(1)
|
33 |
|
34 |
-
return
|
35 |
|
36 |
-
demo = gr.Interface(fn=segment, inputs=["image","text"], outputs="
|
37 |
demo.launch()
|
|
|
5 |
|
6 |
import matplotlib.pyplot as plt
|
7 |
import numpy as np
|
8 |
+
import io
|
9 |
|
10 |
import gradio as gr
|
11 |
|
|
|
15 |
model = CLIPSegForImageSegmentation.from_pretrained("CIDAS/clipseg-rd64-refined")
|
16 |
|
17 |
def visualize_segmentation(image, prompts, preds):
|
18 |
+
fig, ax = plt.subplots(1, len(prompts) + 1, figsize=(3*(len(prompts) + 1), 4))
|
19 |
+
[a.axis('off') for a in ax.flatten()]
|
20 |
+
ax[0].imshow(image)
|
21 |
+
[ax[i+1].imshow(torch.sigmoid(preds[i][0])) for i in range(len(prompts))];
|
22 |
+
[ax[i+1].text(0, -15, prompt) for i, prompt in enumerate(prompts)];
|
23 |
+
|
24 |
+
buf = io.BytesIO()
|
25 |
+
fig.savefig(buf, format='png')
|
26 |
+
plt.close(fig)
|
27 |
+
|
28 |
+
return Image.open(buf)
|
29 |
|
30 |
|
31 |
def segment(img, clases):
|
|
|
32 |
image = Image.fromarray(img, 'RGB')
|
33 |
prompts = clases.split(',')
|
34 |
|
|
|
37 |
outputs = model(**inputs)
|
38 |
preds = outputs.logits.unsqueeze(1)
|
39 |
|
40 |
+
return visualize_segmentation(image, prompts, preds)
|
41 |
|
42 |
+
demo = gr.Interface(fn=segment, inputs=["image","text"], outputs="image")
|
43 |
demo.launch()
|