Spaces:
Runtime error
Runtime error
Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import gradio as gr
|
2 |
+
from PIL import Image
|
3 |
+
from transformers import BlipProcessor, BlipForConditionalGeneration
|
4 |
+
import torch
|
5 |
+
|
6 |
+
# Initialize BLIP model and processor
|
7 |
+
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
8 |
+
processor = BlipProcessor.from_pretrained("Salesforce/blip-image-captioning-base")
|
9 |
+
model = BlipForConditionalGeneration.from_pretrained("Salesforce/blip-image-captioning-base").to(device)
|
10 |
+
|
11 |
+
def caption_image(image):
|
12 |
+
inputs = processor(images=image, return_tensors="pt").to(device)
|
13 |
+
out = model.generate(**inputs)
|
14 |
+
caption = processor.decode(out[0], skip_special_tokens=True)
|
15 |
+
return caption
|
16 |
+
|
17 |
+
def process_image(image):
|
18 |
+
# Convert the input image to PIL Image
|
19 |
+
image = Image.fromarray(image)
|
20 |
+
# Get the caption
|
21 |
+
caption = caption_image(image)
|
22 |
+
return caption
|
23 |
+
|
24 |
+
# Create Gradio Interface
|
25 |
+
interface = gr.Interface(
|
26 |
+
fn=process_image,
|
27 |
+
inputs=gr.inputs.Image(type="numpy", label="Upload Image"),
|
28 |
+
outputs=gr.outputs.Textbox(label="Caption"),
|
29 |
+
title="BLIP Image Captioning",
|
30 |
+
description="Upload an image to get a caption generated by the BLIP model."
|
31 |
+
)
|
32 |
+
|
33 |
+
# Launch the Gradio app
|
34 |
+
if __name__ == "__main__":
|
35 |
+
interface.launch()
|