About Kuih Identifier

© 2022 Mohammad Hafiz bin Ismail

About

Online Kuih Identifier is a web application for identifying Traditional Malay cakes. It is based on the MobileNetV2 Convolutional Neural Network architecture.

The model was trained using TensorFlow with Keras and deployed on the web using the TensorFlow.js library.

Supported Malay Kuih (36 Types)

The model can detect 36 types of traditional Malay kuih. An "unknown" class is also included for non-kuih images.

  • Apam Gula Hangus
  • Apambalik
  • Apom Lenggang
  • Bahulu
  • Bingka
  • Buah Melaka
  • Cara
  • Cara Berlauk
  • Cucur Badak
  • Dendram
  • Karas
  • Karipap
  • Kasturi Kacang Hijau
  • Kek Lapis
  • Keria
  • Ketayap
  • Kuih Bakar Kemboja
  • Kuih Buah Rotan
  • Kuih Cekmek Molek
  • Kuih Sagu
  • Kuih Siput
  • Lompang
  • Loyang
  • Onde Onde
  • Penjaram
  • Popia Basah
  • Popia Goreng
  • Puteri Ayu
  • Putu Piring
  • Roti Jala
  • Samosa
  • Sepit
  • Seri Muka
  • Tepung Gomak
  • Tepung Pelita
  • Tepung Talam

Author

Mohammad Hafiz bin Ismail (mypapit@gmail.com)
For any enquiries, please contact:
deeplearn@uitm.edu.my

Mobile Application

An Android mobile application version is available on the Google Play Store:

Kuih Spotter

Referencing This Work

You may cite this work as:

Ismail, M. H. (2022). Web-based Malay Kuih Identifier Utilizing Convolutional Neural Network. Retrieved from: https://demo.mobilepit.com/ai/kuihweb2/