Monreader

Build a CNN-LSTM architecture to classify sequential images with 93% accuracy.

File Descriptions

• monreader-baseline.ipynb: notebook with data exploration, feature engineering and simple binary image classifier built from scratch
• monreader-optimized.ipynb: notebook with data exploration, feature engineering and binary image classifier built using pretrained models
• monreader-sequence-custom-model.ipynb: notebook with data exploration, feature engineering and the LSTM-CNN architecture from scratch
• monreader-sequence-pretrained-model.ipynb: notebook with data exploration, feature engineering and the LSTM-CNN architecture from pretrained models

Python Libaries Used

• Pandas
• Numpy
• OpenCV
• Sklearn
• Tensorflow
• Matplotlib
• Keras Video Generator