Hey Product Hunt 👋 Forest Fire Classifier started life as a scrappy college project. v2 is the glow-up: a proper web app that can detect fire, smoke, fire+smoke, or no fire from images. What it does: • Upload a forest image → get class + confidence scores in ~500 ms • All inference runs on the server with TensorFlow.js • Images are processed in memory only – nothing is stored Why I built it: I wanted to see how far I could take a Teachable Machine model and turn it into something “portfolio-grade”: proper UX, a11y, rate-limited API, docs, and a clean developer experience for future versions. Who it’s for: • Devs curious about ML in production-ish web apps • Learners looking for a real, inspectable Next.js + TS + TFJS project • Anyone who enjoys seeing old college projects glow up It’s completely free & open source – feel free to fork it, break it, or use the architecture as a reference for your own ML side project. I’d love feedback on: • The UX (especially the classifier flow) • API design & documentation • Anything that feels confusing or slow

Forest Fire Classifier

Instant fire, smoke & no-fire detection from images

Instant fire, smoke & no-fire detection from images