Multi-Class Image Classification Model for Rice Images Identification Using TensorFlow Take 3

Template Credit: Adapted from a template made available by Dr. Jason Brownlee of Machine Learning Mastery.

SUMMARY: This project aims to construct a predictive model using a TensorFlow convolutional neural network (CNN) and document the end-to-end steps using a template. The Rice Images Identification is a multi-class classification situation where we attempt to predict one of several (more than two) possible outcomes.

INTRODUCTION: The dataset owner collected 75,000 pieces of rice grain and created a dataset that classifies the grains into one of the varieties (Arborio, Basmati, Ipsala, Jasmine, Karacadag). The research team applied various preprocessing operations to the rice images and obtained the features. The image size in which each grain of rice is 250 × 250 pixels.

ANALYSIS: The EfficientNetV2S model’s performance achieved an accuracy score of 99.93% after ten epochs using a validation dataset with 20% of total images. After tuning the learning rate, we improved the accuracy rate to 99.94% using the same validation dataset.

CONCLUSION: In this iteration, the TensorFlow EfficientNetV2S CNN model appeared suitable for modeling this dataset.

Dataset Used: Rice Image Dataset

Dataset ML Model: Multi-Class classification with numerical features

Dataset Reference:

One source of potential performance benchmarks:

The HTML formatted report can be found here on GitHub.