Image Classifier

This model uses deep learning and computer vision to predict fruits and vegetables with high accuracy. There are 3 convolutional layers on the convolutional neural network, and an additional 2 fully connected layers. It has 96% accuracy on both training and test predictions.

Key Learnings

• 'categorical_crossentropy' works on one-hot encoded target, while 'sparse_categorical_crossentropy' works on integer target. https://www.kaggle.com/general/197993

References

• Dana Rausch, 2021 - EDA for Image Classification
• StackExchange, 2018 - Exploratory Data Analysis with Image Dataset
• Gaurav Singhal, 2020 - Importing Image Data into NumPy Arrays

Dataset

• Kaggle - Fruits & Vegetables Image Recognition

Documentation & Libraries

• Tensorflow Image Classification
• Dask - Parallel Computing
• Python Imaging Library