Skip to content

My thesis: applied a MATLAB-based ML algorithm to test the effect of spatial frequency on detection advantage of human emotional faces, involving data processing by Python and data modeling

brucelang32/MATLAB-based-machine-learning-algorithm-for-emotion-detection

Repository files navigation

MATLAB-based-machine-learning-algorithm-for-emotion-detection

My thesis: applied a MATLAB-based ML algorithm to test the effect of spatial frequency on detection advantage of human emotional faces, involving data processing by Python and data modeling.

All details are written in my paper, inlcuding the purpose and methond of the whoile project. A Python script of date-preprocessing for experiment are included here as well. Here are some brief description of the algorithm:

  1. The algorithm is based on Viola-Jones object detection framework and is a unsupervised classification
  2. This is a unsupervised classificatio, including custom feature selection and four ways of cross validation:
  • Filter Selection
  • Wrapper Selection
  • Random Selection
  • Pseudo Random Selection
  1. Focus on overall spatial frequency (SF) and localize Histograms of Oriented Gradients (HOG) features.
  2. The modelling approach can further specify which visual features drive these and other behavioural effects related to emotional expressions
  3. We create a Protosc app (see image below) to implement the emotion detection algorithm, you can download here to see how to use it by tutorial.pdf.

image

About

My thesis: applied a MATLAB-based ML algorithm to test the effect of spatial frequency on detection advantage of human emotional faces, involving data processing by Python and data modeling

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages