65 lines
2.3 KiB
Markdown
65 lines
2.3 KiB
Markdown
|
# Tag Scan Mini App
|
||
|
|
|
||
|
|
## Overview
|
||
|
|
|
||
|
|
This project is an AI-powered clothing tag identification and similarity search system. It analyzes clothing tag images, identifies brands using computer vision, and finds similar tags from a database. The system uses advanced AI techniques including image embeddings, text similarity, and (optionally) LLM-based filtering to provide accurate tag matching and recommendations.
|
||
|
|
|
||
|
|
## Features
|
||
|
|
- **Tag Identification:** Uses computer vision to identify clothing tag brands from images
|
||
|
|
- **Text-Based Matching:** Implements TF-IDF and cosine similarity for tag name matching
|
||
|
|
- **Image Similarity Search:** Uses CLIP embeddings and FAISS for visually similar tag images
|
||
|
|
- **LLM Enhancement:** Optional LLM analysis for improved similarity filtering
|
||
|
|
- **Metadata Extraction:** Provides appraisal values, years, and status information for similar tags
|
||
|
|
- **Simple Frontend:** Web UI to upload image URL, toggle LLM, and view results visually
|
||
|
|
|
||
|
|
## Tech Stack
|
||
|
|
- Python, Flask
|
||
|
|
- CLIP (Hugging Face), FAISS, scikit-learn, pandas, numpy
|
||
|
|
- OpenAI (optional, for LLM)
|
||
|
|
- HTML/CSS/JS frontend (Flask template)
|
||
|
|
|
||
|
|
## Setup & Installation
|
||
|
|
|
||
|
|
1. **Clone the Repository**
|
||
|
|
```bash
|
||
|
|
git clone <repository-url>
|
||
|
|
cd ds_task_tag_scan_project/ds_task/backend
|
||
|
|
```
|
||
|
|
2. **Create and Activate Virtual Environment**
|
||
|
|
```bash
|
||
|
|
python3 -m venv ../venv
|
||
|
|
source ../venv/bin/activate
|
||
|
|
```
|
||
|
|
3. **Install Requirements**
|
||
|
|
```bash
|
||
|
|
pip install -r requirements.txt
|
||
|
|
```
|
||
|
|
4. **Set Environment Variables** (if using LLM)
|
||
|
|
```bash
|
||
|
|
export OPENAI_API_KEY=your-openai-key
|
||
|
|
```
|
||
|
|
5. **Run the App**
|
||
|
|
```bash
|
||
|
|
python app.py
|
||
|
|
```
|
||
|
|
(For production, use Docker or a Linux server for stability.)
|
||
|
|
|
||
|
|
## Usage
|
||
|
|
- Go to [http://localhost:8000/](http://localhost:8000/) in your browser
|
||
|
|
- Enter a tag image URL
|
||
|
|
- Toggle "Use LLM Similarity" if desired
|
||
|
|
- Click "Scan Tag" to see results (tag info, similar images, metadata)
|
||
|
|
|
||
|
|
## File Structure
|
||
|
|
- `backend/` - Flask app, ML/DS logic, templates
|
||
|
|
- `data/` - Tag guides, expert and community CSVs
|
||
|
|
- `docs/` - Documentation (this file, API doc)
|
||
|
|
|
||
|
|
## Notes
|
||
|
|
- For best stability, run in a Linux environment or Docker.
|
||
|
|
- On Mac, the app is configured to use only one thread for all ML/numerical libraries.
|
||
|
|
- LLM similarity requires a valid OpenAI API key.
|
||
|
|
|
||
|
|
---
|
||
|
|
|
||
|
|
See `API_Documentation.md` for API details.
|