Feature extraction transfer learning for great results with a small dataset

Transfer learning is leveraging what a state of the art Deep Learning model has already learned

Key features

• Deep Learning for computer vision
• feature extraction for accurate predictions since our dataset is similar to ImageNet from which the TensorflowHub model was trained
• tensorboard callback to track ML training results

Code and resources

• Python version: 3.9
• Packages: zipfile, os, matplotlib, datetime, random, tensorflow, tensorflow_hub, pathlib
• Data source: (https://github.com/DimitrisParaskevopoulos/Transfer-learning-feature-extraction/tree/main/dataset)