get_frames_from_video.ipynb - extract frames to their respective dir from video dataset.
ImageclassifierCNN.ipynb- update required.. tried with CNN using keras/tf and a image as a single data of either normal or abnormal.
suspiciousDetectionFastai.ipynb - Used CNN to do the work using a frame as a single data point and used FastAI to do the task.
After getting frames:
Small_frames_to_feature_map.ipynb - you can store the high level featuremap of the frames in npy file through directory of frames, but just done to small dataset.
big_frames_to_feature_map.ipynb - just bigger version of Small_frames_to_feature_map.ipynb.
Small_RNNImageClassifier.ipynb - Predict from the numpy file of small dataset as dummy dataset of around 15 GB. RNNImageClassifier.ipynb - Predict from the numpy file of dataset using LSTM.
1.csv
2.csv
3.csv
4.csv
final.csv
These CSV files helps us to get frames of the video in correct order easily.