Inspiration

Whenever I post on Social Media, I always get tired to think of creative and unique captions because when you just upload a image, people don't know the story behind it because each image holds a memory aligning with it and a story behind it. So I created this web app which can generate a comprehensive caption related to the Image which you are going to upload.

What it does

It let's you upload an image, and generates a comprehensive caption of that image.

How we built it

I built Pixcribe as a simple web app where you can upload an image and instantly get captions for it. The image is sent to Google Gemini 1.5 Flash, which analyzes the visual content and generates captions.

Challenges we ran into

Handling image uploads and sending them properly to the AI API was a bit tricky at first. We also had to experiment with prompts to get captions that were clear and not too generic.

Accomplishments that we're proud of

We managed to build a clean and simple tool that can generate captions from images almost instantly using AI.

What we learned

We learned how multimodal AI works — how models can understand images and turn them into natural language captions.

What's next for Pixcribe

Next, we want to add features like multiple caption styles, hashtag generation, and maybe support for multiple languages.

Built With

  • google-gemini-1.5-flash-(ai-model)
  • google-generative-ai-api-(api-service)
  • pillow-(image-processing-library)
  • python-(programming-language)
  • streamlit-(web-framework/ui)
  • streamlit-platform-(application-interface)
Share this project:

Updates