Fruits Classification using Convolutional Neural Networks (CNN)
This project is an introduction to deep learning via CNN. The objective is to build and optimise a model to classify fruit images using Tensorflow and Keras. Image augmentation was one aspect to optimise the model. Individual performance of different architectures were tracked using metrics such as accuracy, validation loss and early stopping.
If .ipynb file fails to render, you can view it here
Python Libraries Used [Python Version: 3.10.11]
- Pandas
- Numpy
- Matplotlib
- Seaborn
- Tensorflow
Features
- Training and evaluation of CNN on single-fruit images
- Visualisation of training and validation accuracy and loss
- Random testing of model's prediction on fruits in dataset
- Comparison of models with and without data augmentation
Results
- Can be found in my report showcasing my analysis findings.
Built With
- jupyter-notebook
