🌭 Hotdog or Not — Next.js + Hugging Face

A simple image classification project built with Next.js (TypeScript) that tells whether an uploaded image contains a hotdog or not using a pre-trained Vision Transformer (ViT) model from Hugging Face.

Tech Stack

• Next.js (App Router)
• TypeScript
• pnpm
• Hugging Face Inference API
• google/vit-base-patch16-384 image classification model

🧠 How it works

• User uploads an image
• Image is sent to a Next.js API route
• API calls Hugging Face Inference API
• Model returns image labels
• App checks for “hotdog” and shows the result

Notes

• The model runs on Hugging Face (no local ML setup needed)
• This is a basic demo project, not a fine-tuned classifier
• Accuracy depends on the general image classification model