Data-driven approach to learning salience models of indoor landmarks by using genetic programming

By Xuke, Lei Ding, Jianga Shang, Hongchao Fan, Tessio Novack, Alexey Noskov, Alexander

Introduction

This paper addresses the problem of determining the most salient landmark from several candidates at decision points in landmark-based way-finding. Our method significantly outperforms existing methods and achieves around 76% precision. This accuracy rate is considerably higher than the ones achieved by conventional linear models. To learn more, please refer to our paper(https://www.tandfonline.com/doi/full/10.1080/17538947.2019.1701109)

Code

Our code is under "code/" folder. The code is implemented by Matlab R2018a. The algorithm is based on GPTIPS Symbolic Machine Learning Platform for MATLAB (https://sites.google.com/site/gptips4matlab/home)

Data

Our data is under "data/" folder. Lists of "data/scencePic" folder are navigation scence images. The sheet of "featuresAttributes" in landmarkDataSets.xlsx file is landmarks attributes value, and the sheet of "questionnairesResult" is questionnaire results from volunteers.

Usage

The main function is Main_fold_5cross_validation.m, you could run the main function.

Contact

Comments, queries and bug reports are appreciated. Email: mylovedinglei@gmail.com