Derek i am Facing Issues while running the code Need some help in it , I have placed the sample images in data folder . Do we need any files for landmark and detected folder ? about audetect

[derek-schultz]

@shonivee

haarcascade_frontalface_alt.xml

You have to get the landmarks and images from the Cohn-Kanade Database. In the CK+ directory there is a Landmarks.zip file that contains all of the landmarks text files sorted by session (example: S005_001_00000001_landmarks.txt).

Let me know if you are able to get it sorted out.

[derek-schultz]

First you have to run the training.

audetect-interactive.py train --features
audetect-interactive.py train --action_units

Then you should be able to run a prediction on one of the sequences like this
audetect-interactive.py detect --sequence data/images/S130/001

[derek-schultz]

Sorry, I probably didn't document this enough. The data directory should contain haarcascade_frontalface_alt.xml, and directories called images and landmarks. The images directory should contain the sessions (S014, S022, etc) and the sessions should contain PNG images.
The landmarks directory should also contain sessions and those sessions should contain the landmarks.txt files.

[Sho-s01]

can you send the link for landmarks.txt file and haarcascade_frontalface_alt.xml ??

[Sho-s01]

@derek-schultz
thank you so much for the

reply.I have included the landmarks and images in the data folder.but then i am getting the following error.

[derek-schultz]

@shonivee I just pushed what might be a fix. Please pull the latest code and try again?

[kowsalya1997]

@derek-schultz

I am getting the following issue.Please fix this

[derek-schultz]

@kowsalya1997 Is this what happens when you are running audetect-interactive.py detect? Did you first run both audetect-interactive.py train --features and audetect-interactive.py train --action_units?

[nandy26]

hello @derek-schultz ,

can you tell me the correct order to run the code??
Thanks in advance

[nandy26]

@derek-schultz

Please upload the video demonstration of how train the features and action units ??

[kowsalya1997]

@derek-schultz

I am getting the following errors .please fix this .
on running audetect-interactive.py train --features

on running audetect-interactive.py train --action_units,

on running audetect-interactive.py detect --sequence data/images/S130/001,

Thanks.

[kowsalya1997]

@derek-schultz

please fix the above issue.Awaiting for your reply...