README.md

metadata

title: Sustainable Content Moderation
emoji: 🌍
colorFrom: green
colorTo: blue
sdk: gradio
sdk_version: 5.16.0
app_file: app.py
pinned: false

Multi-Model multi-agent Analysis System

A multi-agent system for analyzing text and images using state-of-the-art AI models.

Overview

This application uses multiple AI agents to:

• Process user queries (check if positive or negative)
• Analyze images that the user uploads
• Search relevant technical context (analyzes if the text and images are matching and related)
• Generate comprehensive analysis reports (see if text is positive/ negative and image is positive/ negative. subsequently check if text and images are coorelated and perform analysis)

Usage

1. Enter your query
2. (Optional) Add any specific constraints
3. Upload equipment images (supported formats: JPG, JPEG, PNG)
4. Select number of top results to show
5. Choose report format (summary/detailed)
6. Click "Analyze" to process

Technical Details

• Built with Gradio 5.16.0
• Uses multi-model approach:
  • Lightweight models for initial processing
  • Advanced models for detailed analysis
• Implements memory-efficient processing
• Supports batch image processing

Limitations

• Maximum image size: 5MB
• Maximum resolution: 2048x2048
• Maximum images per request: 10
• Query length limit: 500 characters

Models Used

• Image Captioning: BLIP and BLIP-2
• Image Classification: ResNet and ViT
• Text Processing: LaMini-Flan-T5

Requirements

• Python 3.8+
• See requirements.txt for full dependencies

Local Development

pip install -r requirements.txt python app.py

Deployment

This app is optimized for HuggingFace Spaces deployment.

Error Handling

• Input validation with clear error messages
• Resource monitoring and automatic cleanup
• Graceful error recovery

Credits

• Built using HuggingFace's model hub
• Powered by Gradio interface
• Uses Wikipedia API for context gathering

License

MIT License

Support

For issues or questions, please open a GitHub issue.