DermaDetect AI

Inspiration

I was inspired by how many people ignore or misdiagnose skin conditions, sometimes leading to serious health issues. I wanted to create a tool that could bring early detection to everyone using just a phone.

What I Built

I developed DermDetect, an AI-powered app that allows users to upload a photo of a skin condition. The AI analyzes the image and predicts possible skin conditions with a risk score and recommendations for further action.

How I Built It

Used Python and TensorFlow for image recognition.

Trained the model on public skin disease datasets.

Built a simple React frontend where users can upload images and see results.

Deployed a prototype quickly for hackathon demo using Streamlit/Firebase.

Challenges I Faced

Finding clean, labeled datasets suitable for AI training.

Ensuring the model gives accurate predictions on different skin tones and lighting conditions.

Making the app fast and user-friendly for a live demo.

What I Learned

AI can solve real-world health problems if used carefully.

Building a functional prototype in a short time requires prioritizing key features.

Presenting medical AI responsibly is important – the app gives guidance, not diagnoses.

I want to do the Open Track BTW

Built With

• datasets
• firebase-for-database-and-hosting
• python
• react-for-the-frontend
• streamlit-for-the-demo
• tensorflow-for-ai-image-recognition