Eye movement event detection using threshold based(Velocity threshold and Dispersion threshold), machine learning algorithms(Random forest) and deep learning algorithms(CNN). In this project we evaluated different event detection algorithms using eye tracking data. the parameter used to represent the data is velocity. Whcih means sequwnce of x,y points are converted in to horizontal and vertical velocity. Conersion is done by claculating point to point velocities by using point to point distance and time. We use velocity for all classifiers in order to make fair evaluation and comparisoin.
tobiasweede / event_detection Goto Github PK
View Code? Open in Web Editor NEWEye movement event detection using threshold based(Velocity threshold and Dispersion threshold), machine learning algorithms(Random forest and KNN) and deep learning algorithms(CNN, LSTM and combination of LSTM and 2DCNN).