File size: 2,855 Bytes
632e4ae
 
 
 
1f13a56
632e4ae
 
 
fe1a29a
632e4ae
 
1f13a56
 
632e4ae
85cbea1
632e4ae
1f13a56
 
 
 
 
 
 
fe1a29a
 
 
 
 
 
 
1f13a56
fe1a29a
1f13a56
 
fe1a29a
1f13a56
 
fe1a29a
1f13a56
 
fe1a29a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1f13a56
fe1a29a
1f13a56
fe1a29a
 
 
 
 
 
 
1f13a56
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
import os
import sys
src_directory = os.path.abspath(os.path.join(os.path.dirname(__file__), "../..", "src"))
sys.path.append(src_directory)
from transformers import AutoProcessor, CLIPModel
import streamlit as st
from utils import logger
from database import pinecone_index
from PIL import Image

logger = logger.get_logger()
model = CLIPModel.from_pretrained("openai/clip-vit-base-patch32")
processor = AutoProcessor.from_pretrained("openai/clip-vit-base-patch32")

PINECONE_INDEX =  pinecone_index.create_index()

def search_by_text(query_text, index):
    inputs = processor(text=query_text, return_tensors="pt")
    text_features = model.get_text_features(**inputs)
    query_vector = text_features.detach().cpu().numpy().flatten().tolist()
    results = index.query(vector=query_vector, top_k=10, include_metadata=True, namespace="image-search-dataset")
    return results

def search_by_image(image, index):
    inputs = processor(images=image, return_tensors="pt")
    image_features = model.get_image_features(**inputs)
    query_vector = image_features.detach().cpu().numpy().flatten().tolist()
    results = index.query(vector=query_vector, top_k=5, include_metadata=True, namespace="image-search-dataset")
    return results

def main():
    st.set_page_config(page_title="Clip Search", layout="wide")
    st.title("📸Image Search with Pinecone and CLIP")
    option = st.selectbox("Choose Input Type", ["Text", "Image Upload"])
    
    if option == "Text":
        user_text = st.text_input("Enter your search text", placeholder = "for eg: dogs or cat etc..")
        
        if st.button("Search"):
            results = search_by_text(user_text, PINECONE_INDEX)
            
            columns = st.columns(2)

            for idx, match in enumerate(results['matches']):
                with columns[idx % 2]:
                    st.image(
                        match['metadata']['url'], 
                        caption=f"Match: {match['metadata']['photo_id']}", 
                        width=500
                            )

    elif option == "Image Upload":
        uploaded_file = st.file_uploader("Upload an image", type=["jpg", "png", "jpeg"])
        if uploaded_file is not None:
            image = Image.open(uploaded_file)
            st.image(image, caption="Uploaded Image")
            if st.button("Search by Image"):
                results = search_by_image(image, PINECONE_INDEX)

                columns = st.columns(2)

                for idx, match in enumerate(results['matches']):
                    with columns[idx % 2]:
                        st.image(
                            match['metadata']['url'], 
                            caption=f"Match: {match['metadata']['photo_id']}", 
                            width=500
                                )
 
if __name__ == "__main__":
    main()