An AI-powered application that generates engaging Instagram captions from images.
The system combines computer vision, large language models, and semantic ranking to create stylish captions with emojis, multilingual translations, and engagement scoring.
Built using Streamlit, BLIP Vision Models, Groq Llama LLMs, and CLIP embeddings.
- Generate Instagram captions from images
- Multiple caption styles (Instagram, Funny, Savage, Aesthetic, Motivational, Poetic)
- Emoji caption generation
- Punjabi translation support
- Viral score prediction (engagement estimate)
- Caption ranking using CLIP embeddings
- Automatic hashtag generation
- Caption rating and dataset collection
- Download generated captions
- Streamlit web interface
Image Upload ↓ BLIP Vision Model ↓ Image Description ↓ Groq Llama LLM ↓ Caption Package ↓ CLIP Ranking ↓ Streamlit UI
The system combines:
- BLIP → Image understanding
- Groq Llama → Caption generation
- CLIP → Caption ranking
- Streamlit → Interactive UI
| Component | Technology |
|---|---|
| Frontend | Streamlit |
| Vision Model | BLIP (HuggingFace Transformers) |
| LLM | Groq Llama |
| Semantic Ranking | CLIP (Sentence Transformers) |
| Language | Python |
| Deployment | Streamlit Cloud |
Clone the repository:
git clone https://github.com/rachit23tech/ai-instagram-caption-generator.git
cd ai-instagram-caption-generator
##Install dependencies:
pip install -r requirements.txt