Reko

Simple Python scripts for supporting face based user verification.

1. Obtain verbal instructions using Google API - speechrecognition to get command. (Improvement required) (TODO - use Amazon Transcribe instead)
2. Control webcam to take image using OpenCV 2.
3. Face recognition with Amazon Rekognition.
4. Verbal feedback (Text-to-Speech) using Amazon Polly.

• LICENSE
• Release Notes

Build

Python 3.7, 3.8:

 virtualenv env
 env\Scripts\activate
 pip install -i https://pypi.python.org/pypi -e .

Run

1. Get Help from command line

   reko --help
   

2. List all collections from command line

   reko --profile AWS_PROFILE_NAME --collections
   

3. List all faces in a collection

   reko --profile AWS_PROFILE_NAME --collection_id COLLECTION_ID --faces
   

4. Sign-up from command line. If collection does not exist, it will be created

   reko --profile AWS_PROFILE_NAME --collection_id COLLECTION_ID --signup NAME
   

5. Sign-in from command line, with verbal feedback

   reko --profile AWS_PROFILE_NAME --collection_id COLLECTION_ID --signin NAME --audio_on
   

6. Delete a collection from command line

   reko --profile AWS_PROFILE_NAME --collection_id COLLECTION_ID --delete_collection
   

7. Sign-in using microphone

   reko --profile AWS_PROFILE_NAME --collection_id COLLECTION_ID --audio_on --listen_on
   

8. Keep watching (at the specified interval) and sign-in when possible

   reko --profile AWS_PROFILE_NAME --collection_id COLLECTION_ID -audio_on --watch_on INTERVAL_SEC