Spaces:
Running
Running
Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,76 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import requests
|
2 |
+
from PIL import Image
|
3 |
+
from io import BytesIO
|
4 |
+
import torch
|
5 |
+
from torchvision import transforms
|
6 |
+
from transformers import AutoModelForImageClassification, AutoConfig
|
7 |
+
import gradio as gr
|
8 |
+
|
9 |
+
model_id = "thelabel/240903-image-tagging"
|
10 |
+
config = AutoConfig.from_pretrained(model_id)
|
11 |
+
model = AutoModelForImageClassification.from_pretrained(model_id)
|
12 |
+
model.eval()
|
13 |
+
|
14 |
+
# Standard ViT image transforms
|
15 |
+
image_transform = transforms.Compose([
|
16 |
+
transforms.Resize((224, 224)),
|
17 |
+
transforms.ToTensor(),
|
18 |
+
transforms.Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5]),
|
19 |
+
])
|
20 |
+
|
21 |
+
def load_image_from_url(url):
|
22 |
+
try:
|
23 |
+
response = requests.get(url, timeout=10)
|
24 |
+
response.raise_for_status()
|
25 |
+
return Image.open(BytesIO(response.content)).convert("RGB")
|
26 |
+
except Exception as e:
|
27 |
+
return None
|
28 |
+
|
29 |
+
def predict_tags(image_url, threshold=0.5):
|
30 |
+
image = load_image_from_url(image_url)
|
31 |
+
if image is None:
|
32 |
+
return [], "Could not load image from the provided URL."
|
33 |
+
|
34 |
+
image_tensor = image_transform(image).unsqueeze(0)
|
35 |
+
with torch.no_grad():
|
36 |
+
logits = model(image_tensor).logits
|
37 |
+
probs = torch.sigmoid(logits).squeeze()
|
38 |
+
|
39 |
+
results = [
|
40 |
+
(config.idx_to_label[str(i)], float(probs[i]))
|
41 |
+
for i in range(len(probs))
|
42 |
+
if probs[i] >= threshold
|
43 |
+
]
|
44 |
+
results.sort(key=lambda x: x[1], reverse=True)
|
45 |
+
return results, None
|
46 |
+
|
47 |
+
def gradio_predict(url, threshold):
|
48 |
+
tags, error = predict_tags(url, threshold)
|
49 |
+
if error:
|
50 |
+
return error, None
|
51 |
+
return "\n".join([f"{tag}: {score:.2f}" for tag, score in tags]), url
|
52 |
+
|
53 |
+
demo = gr.Interface(
|
54 |
+
fn=gradio_predict,
|
55 |
+
inputs=[
|
56 |
+
gr.Textbox(label="Image URL", value="https://d2q1sfov6ca7my.cloudfront.net/eyJidWNrZXQiOiJoaWNjdXAtaW1hZ2UtaG9zdGluZyIsImtleSI6ImhpY2N1cC1wcm9kdWN0cy9GQVFZTFkyNzFGLmpwZWciLCJlZGl0cyI6eyJyZXNpemUiOnsid2lkdGgiOjI1NjAsImhlaWdodCI6Mzg0MCwiZml0IjoiY292ZXIifX19?v=1748968367"),
|
57 |
+
gr.Slider(0, 1, value=0.5, step=0.01, label="Threshold"),
|
58 |
+
],
|
59 |
+
outputs=[
|
60 |
+
gr.Textbox(label="Tags"),
|
61 |
+
gr.Image(label="Preview", type="url"),
|
62 |
+
],
|
63 |
+
title="Image Tagging with ViT",
|
64 |
+
description="Paste an image URL and get predicted tags using thelabel/240903-image-tagging model.",
|
65 |
+
examples=[
|
66 |
+
[
|
67 |
+
"https://d2q1sfov6ca7my.cloudfront.net/eyJidWNrZXQiOiJoaWNjdXAtaW1hZ2UtaG9zdGluZyIsImtleSI6ImhpY2N1cC1wcm9kdWN0cy9GQVFZTFkyNzFGLmpwZWciLCJlZGl0cyI6eyJyZXNpemUiOnsid2lkdGgiOjI1NjAsImhlaWdodCI6Mzg0MCwiZml0IjoiY292ZXIifX19?v=1748968367", 0.5
|
68 |
+
],
|
69 |
+
[
|
70 |
+
"https://d2q1sfov6ca7my.cloudfront.net/eyJidWNrZXQiOiJoaWNjdXAtaW1hZ2UtaG9zdGluZyIsImtleSI6ImhpY2N1cC1wcm9kdWN0cy9ON01aQkpUMDlFLmpwZWciLCJlZGl0cyI6eyJyZXNpemUiOnsid2lkdGgiOjI1NjAsImhlaWdodCI6Mzg0MCwiZml0IjoiY292ZXIifX19?v=1748968367", 0.5
|
71 |
+
]
|
72 |
+
]
|
73 |
+
)
|
74 |
+
|
75 |
+
if __name__ == "__main__":
|
76 |
+
demo.launch()
|