Skip to content

Latest commit

 

History

History
66 lines (57 loc) · 2.43 KB

README.md

File metadata and controls

66 lines (57 loc) · 2.43 KB

RempahRasaML

Introduction

Repository contains image classification models with 31 classes. The models are trained using custom CNN methods and transfer learning to recognize Indonesian spices images. Five architectures that we use for transfer learning includes: InceptionV3, Xception, DenseNet121, EfficientNetV2B0, and MobileNetV3. After the training, we find out that the model with the highest accuracy is DenseNet121 architecture with 94.37% accuracy on test dataset.

Spices Classification Model

Dataset

There are total 6510 data and 31 classes of indonesian spices images including:

  1. adas
  2. andaliman
  3. asam jawa
  4. bawang bombai
  5. bawang merah
  6. bawang putih
  7. biji ketumbar
  8. bukan rempah
  9. bunga lawang
  10. cengkeh
  11. daun jeruk
  12. daun kemangi
  13. daun ketumbar
  14. daun salam
  15. jahe
  16. jinten
  17. kapulaga
  18. kayu manis
  19. kayu secang
  20. kemiri
  21. kemukus
  22. kencur
  23. kluwek
  24. kunyit
  25. lada
  26. lengkuas
  27. pala
  28. saffron
  29. serai
  30. vanili
  31. wijen

Model Architecture

transfer_learning_model_custom (1)

Training and Evaluation

Pretraining

  • Before training, we have splitted the data into 70% for training, 15% for validation, and 15% for test.
  • Visualization some of the data image

Training

  • We have trained the model and after 50 epochs, we obtained loss: 0.1228 and accuracy 96.44% for training. For validation, the val_loss is 0.3062 with 93.35% accuracy. image

Evaluation

  • We also have evaluated the model on test set, and achieved loss 0.23866 with 94.37% accuracy. image
  • Visualization of data prediction image
  • Confusion matrix Confusion Matrix

Spices Classification model file (.h5 and tfjs for deployment)

https://drive.google.com/drive/folders/1axCQ-LzEvcT-2hm-bzMSf215pR7z8iWe?usp=sharing