Easily convert RGB video data (e.g. .avi) to the TensorFlow tfrecords file format for training e.g. a NN in TensorFlow. This implementation allows to limit the number of frames per video to be stored in the tfrecords.
In this PR, I changed the NumPy array type from uint32 to uint8 to reduce the memory and storage consumption. I also added a rule to avoid jpeg_encoding in the case of using dense optical flow, once the jpeg only has 3 channels, while with the flow the images have 4 channels.
In this PR, I changed the NumPy array type from uint32 to uint8 to reduce the memory and storage consumption. I also added a rule to avoid jpeg_encoding in the case of using dense optical flow, once the jpeg only has 3 channels, while with the flow the images have 4 channels.