A desktop GUI application that enables the training, testing, evaluation of a custom trained Image Classifier using State-of-Art pre-trained CNN models.
Train and test a custom image classifier with only a few button clicks.
Demo : https://youtu.be/moriDlpUWUU
Instalation & use tutorial: https://youtu.be/_8V9j0sGSDQ
Code explanation and walkthough: https://youtu.be/neviJTiJbWQ
All the CNN models and operations are downloaded form TensorFlow hub: https://tfhub.dev/
To do:
- Compress into an exicutable standalone program
- Include a section within the window that shows the output of the terminal
- Reformat the output of the confusion matrix to include class names and consider performance in percentages rather than numbers
This project is based on the development of an older project CNN-Experiment-Runner: https://github.com/KhalidAlnujaidi/CNN-Experiment-Runner
xyz