Continuous-time Prediction of Dimensional Behavior/Affect

 

DOWNLOAD MATLAB CODE for the predictive framework that addresses continuous-time prediction of dimensional affect or behavior developed for the SEWA project. The framework explicitly model temporal dynamics of affective or social behavior displays  based on Linear Time-Invariant (LTI) system learning and can be used to predict future values of affect or behavior based on past observations from the same sequence. The package includes also a demo script that shows the application of the proposed framework for Input-Output LTI system learning used for conflict intensity prediction on a video from the CONFER DatabaseNote that this code can be run only on 64-bit machines with Windows, Linux or Mac OS operating systems.

 

Code released as is **for research purposes only**.

 

 

contact: christos.georgakis@imperial.ac.uk