FlavorSnap

FlavorSnap is a food image classification web app powered by deep learning. Simply upload a picture of a local dish, and the model will tell you what it is. The app uses a fine-tuned ResNet18 model to recognize and classify various food types such as Akara, Bread, Egusi, Moi Moi, Rice and Stew, and Yam (with the option to add more food variety).

Features

• Image Upload & Preview
  Instantly preview your uploaded image in the sidebar.

• AI-Powered Classification
  Classifies the image using a ResNet18 model trained on local food images.

• Automatic Image Organization
  Uploaded images are saved into their predicted class folders.

• Web UI
  Intuitive interface built with html css javascript and typescript.

• backend

• using nextjs.

---

Model Info

• frontend: `html,css javascript,typescript'
• Backbone: ResNet18
• Framework: PyTorch
• Custom classes:
  • Akara
  • Bread
  • Egusi
  • Moi Moi
  • Rice and Stew
  • Yam

The link to the telegram group for issues communication https://t.me/+Tf3Ll4oRiGk5ZTM0