app-seg.py

import streamlit as st
import open_clip
import torch
import requests
from PIL import Image
from io import BytesIO
import time
import json
import numpy as np
import cv2
import chromadb
from transformers import pipeline
import torch.nn as nn
import matplotlib.pyplot as plt

# Load CLIP model and tokenizer
@st.cache_resource
def load_clip_model():
    model, preprocess_train, preprocess_val = open_clip.create_model_and_transforms('hf-hub:Marqo/marqo-fashionSigLIP')
    tokenizer = open_clip.get_tokenizer('hf-hub:Marqo/marqo-fashionSigLIP')
    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
    model.to(device)
    return model, preprocess_val, tokenizer, device

clip_model, preprocess_val, tokenizer, device = load_clip_model()

# Load Clothing Segmentation model
@st.cache_resource
def load_segmentation_model():
    return pipeline(model="mattmdjaga/segformer_b2_clothes")

segmenter = load_segmentation_model()

# Load ChromaDB
@st.cache_resource
def load_chromadb():
    client = chromadb.PersistentClient(path="./clothesDB")
    collection = client.get_collection(name="clothes_items_ver3")
    return collection

collection = load_chromadb()

# Helper functions
def load_image_from_url(url, max_retries=3):
    for attempt in range(max_retries):
        try:
            response = requests.get(url, timeout=10)
            response.raise_for_status()
            img = Image.open(BytesIO(response.content)).convert('RGB')
            return img
        except (requests.RequestException, Image.UnidentifiedImageError) as e:
            if attempt < max_retries - 1:
                time.sleep(1)
            else:
                return None

def get_image_embedding(image):
    image_tensor = preprocess_val(image).unsqueeze(0).to(device)
    with torch.no_grad():
        image_features = clip_model.encode_image(image_tensor)
        image_features /= image_features.norm(dim=-1, keepdim=True)
    return image_features.cpu().numpy()

def get_text_embedding(text):
    text_tokens = tokenizer([text]).to(device)
    with torch.no_grad():
        text_features = clip_model.encode_text(text_tokens)
        text_features /= text_features.norm(dim=-1, keepdim=True)
    return text_features.cpu().numpy()

def find_similar_images(query_embedding, collection, top_k=5):
    database_embeddings = np.array(collection.get(include=['embeddings'])['embeddings'])
    similarities = np.dot(database_embeddings, query_embedding.T).squeeze()
    top_indices = np.argsort(similarities)[::-1][:top_k]
    
    all_data = collection.get(include=['metadatas'])['metadatas']
    
    results = [
        {'info': all_data[idx], 'similarity': similarities[idx]}
        for idx in top_indices
    ]
    return results

def segment_clothing(img, clothes=["Hat", "Upper-clothes", "Skirt", "Pants", "Dress", "Belt", "Left-shoe", "Right-shoe", "Scarf"]):
    segments = segmenter(img)
    mask_list = []
    for s in segments:
        if s['label'] in clothes:
            mask_list.append(s['mask'])
    
    if mask_list:
        final_mask = np.array(mask_list[0])
        for mask in mask_list[1:]:
            current_mask = np.array(mask)
            final_mask = final_mask + current_mask
        
        final_mask = Image.fromarray(final_mask.astype('uint8') * 255)
        img = img.convert("RGBA")
        img.putalpha(final_mask)
    
    return img, segments

# Streamlit app
st.title("Advanced Fashion Search App")

# Initialize session state
if 'step' not in st.session_state:
    st.session_state.step = 'input'
if 'query_image_url' not in st.session_state:
    st.session_state.query_image_url = ''
if 'segmentations' not in st.session_state:
    st.session_state.segmentations = []
if 'selected_category' not in st.session_state:
    st.session_state.selected_category = None

# Step-by-step processing
if st.session_state.step == 'input':
    st.session_state.query_image_url = st.text_input("Enter image URL:", st.session_state.query_image_url)
    if st.button("Segment Clothing"):
        if st.session_state.query_image_url:
            query_image = load_image_from_url(st.session_state.query_image_url)
            if query_image is not None:
                st.session_state.query_image = query_image
                segmented_image, st.session_state.segmentations = segment_clothing(query_image)
                st.session_state.segmented_image = segmented_image
                if st.session_state.segmentations:
                    st.session_state.step = 'select_category'
                else:
                    st.warning("No clothing items segmented in the image.")
            else:
                st.error("Failed to load the image. Please try another URL.")
        else:
            st.warning("Please enter an image URL.")

elif st.session_state.step == 'select_category':
    col1, col2 = st.columns(2)
    with col1:
        st.image(st.session_state.query_image, caption="Original Image", use_column_width=True)
    with col2:
        st.image(st.session_state.segmented_image, caption="Segmented Image", use_column_width=True)

    st.subheader("Segmented Clothing Items:")
    
    options = list(set(s['label'] for s in st.session_state.segmentations))
    selected_option = st.selectbox("Select a category to search:", options)
    
    if st.button("Search Similar Items"):
        st.session_state.selected_category = selected_option
        st.session_state.step = 'show_results'

elif st.session_state.step == 'show_results':
    st.image(st.session_state.query_image, caption="Query Image", use_column_width=True)
    st.image(st.session_state.segmented_image, caption="Segmented Image", use_column_width=True)
    
    selected_segment = next(s for s in st.session_state.segmentations if s['label'] == st.session_state.selected_category)
    mask = np.array(selected_segment['mask'])
    masked_image = Image.fromarray((np.array(st.session_state.query_image) * mask[:,:,None]).astype('uint8'))
    
    st.image(masked_image, caption=f"Selected Category: {st.session_state.selected_category}", use_column_width=True)
    
    query_embedding = get_image_embedding(masked_image)
    similar_images = find_similar_images(query_embedding, collection)
    
    st.subheader("Similar Items:")
    for img in similar_images:
        col1, col2 = st.columns(2)
        with col1:
            st.image(img['info']['image_url'], use_column_width=True)
        with col2:
            st.write(f"Name: {img['info']['name']}")
            st.write(f"Brand: {img['info']['brand']}")
            category = img['info'].get('category')
            if category:  
                st.write(f"Category: {category}")
            st.write(f"Price: {img['info']['price']}")
            st.write(f"Discount: {img['info']['discount']}%")
            st.write(f"Similarity: {img['similarity']:.2f}")
    
    if st.button("Start New Search"):
        st.session_state.step = 'input'
        st.session_state.query_image_url = ''
        st.session_state.segmentations = []
        st.session_state.selected_category = None

# Text search (optional, you can keep or remove this part)
st.sidebar.title("Text Search")
query_text = st.sidebar.text_input("Enter search text:")
if st.sidebar.button("Search by Text"):
    if query_text:
        text_embedding = get_text_embedding(query_text)
        similar_images = find_similar_images(text_embedding, collection)
        st.sidebar.subheader("Similar Items:")
        for img in similar_images:
            st.sidebar.image(img['info']['image_url'], use_column_width=True)
            st.sidebar.write(f"Name: {img['info']['name']}")
            st.sidebar.write(f"Brand: {img['info']['brand']}")
            category = img['info'].get('category')
            if category:
                st.sidebar.write(f"Category: {category}")
            st.sidebar.write(f"Price: {img['info']['price']}")
            st.sidebar.write(f"Discount: {img['info']['discount']}%")
            st.sidebar.write(f"Similarity: {img['similarity']:.2f}")
    else:
        st.sidebar.warning("Please enter a search text.")

# Display ChromaDB vacuum message
st.sidebar.warning("If you've upgraded ChromaDB from a version below 0.6, you may benefit from vacuuming your database")