Summer Research Project by Team (Arjan singh bal, Mathew Sachin, Vipul Barodiya) under mentorship of Prof. Aditya Trivedi and Dr. Sunil Kumar at ABV-IIITM Gwalior.

This is an work focused on implementing ( and experimenting new Models) parts of methods suggested by R. Vera et.al. in their paper Comparative Analysis and Fusion of Spatio-Temporal Information for Footstep Recognition", in IEEE Pattern Analysis and Machine Intelligence, 2012 (DOI: 10.1109/TPAMI.2012.164) ".

The basic idea of project is to capture and finding features of the pressure signature of a human being to classifying them on client vs impostor classes based in various scenarios (Datasets of different size - Lab, Airport security, Home).

The Database SFOOTBD was used for research and academic purpose with permissions from the Biometrics and Data Pattern Analytics - BiDA Lab.

Frame- 1

Frame-2

Frame-3

Global GRF, upper contour, lower contour and various other features were experimented and thier effect on the overall resukts were analysed.

Global GRF Profile -

Upper contour -

Lower contour -

Standard deviation -

Signal Average -

Using XGBoost over SVM had significant improvements in the classifying Spatially processed signatures which also contributed to overal classification after ensembling.

##Quick-start Download SFootBD.zip and extract in this folder.

SFootBD
  - IndexFiles
  - SFootBD
BTP.ipynb

[PAMI2012] R. Vera-Rodriguez, J. Mason, J. Fierrez and J. Ortega-Garcia, "Comparative Analysis and Fusion of Spatio-Temporal Information for Footstep Recognition", in IEEE Pattern Analysis and Machine Intelligence, 2012 (DOI: 10.1109/TPAMI.2012.164).

[IET2011] R. Vera-Rodriguez, J. Mason, J. Fierrez and J. Ortega-Garcia, "Analysis of Spatial Domain Information for Footstep Recognition", IET Computer Vision, 5, 380-388, 2011 (DOI: 10.1049/iet-cvi.2010.0189).