rhodricusack / aws_video

Code used to run Chouinard et al

2020-08-05

rekognition_all_stages.py

Now single entry point to run rekognition stages.

2018-02-26

run_video_rekognition.py

Picks up mp4 files from S3 and runs rekognition on them

process_rekognition_video.py

Retrieves completion messages from SQS and matches behavioural data with automatically coded data
Saves results in pickle files in s3://infantrekognition/coding
Creates annotated videos, which it saves in s3://infantrekognition/annotated

post_rekognition_ml.py

Loads coding pickle files and extracts creates Pandas dataframe containing columns that are likely to be useful for ML