contextual information in action recognition

Hi, i looking for a code that can recognise a human action more accurately by using the object he/she is holding. is there any code available for download ? dataset : youtube action dataset [UCF sport dataset ]

Answer 1

[Image Analyst]

See Feature Matching in the Computer Vision System Toolbox.

Combine feature matching of the thing she is holding with gesture recognition. You can find papers on gesture recognition here: http://www.visionbib.com/bibliography/contentspeople.html#Face%20Recognition,%20Detection,%20Tracking,%20Gesture%20Recognition,%20Fingerprints,%20Biometrics

[Kelly Howard]

thank you for replying. But i am looking for code from the paper like "Action Spotting and Recognition Based on Spatiotemporal Orientation Analysis",

[Walter Roberson]

If you are looking for code to a specific paper, write to the corresponding author of the paper and ask for it.

http://www.seas.upenn.edu/~derpanis/