🧠 Inspiration We wanted to explore how AI can describe visuals the way humans do. Inspired by accessibility use cases and automation in content creation, we built an image captioning tool using Python and transformers.
💡 What it does It automatically generates human-like captions for any input image using the BLIP transformer model.
🔧 How we built it Python
Hugging Face Transformers (Salesforce/blip-image-captioning-base)
PIL for image handling
🚧 Challenges we ran into Handling image formats
Installing large transformer models locally
Making captions concise and readable
🏆 Accomplishments we're proud of Successfully implemented an end-to-end AI captioning flow
Learned how vision-language models work
📚 What we learned Working with pre-trained transformer models
Basics of image preprocessing and token decoding
Handling real-world AI applications with minimal code