TOWARDS THE ADAPTIVE IDENTIFICATION OF WALKERS : AUTOMATED FEATURE SELECTION OF FOOTSTEPS USING DISTINCTION-SENSITIVE LVQ

We applied a method called Distinction-Sensitive Learning Vector Quantization (DSLVQ) to the classification of footsteps. The measurements were made by a pressure-sensitive floor, which is part of the smart sensing living room in our research laboratory. The aim is to identify walkers based on their single footsteps. DSLVQ is an extended version of Learning Vector Quantization (LVQ), and it can be used for automated feature scaling and selection during the trai ning of an LVQ codebook. The method shows improvements in the classification accuracies compared to a standard LVQ. In addition, due to its capability of automated input prunin g, discarding the non-informative features, it was able to det ect automatically the most significant features from a large set of features. This is important in an adaptive identification system, where the informative features might change.

[1]  Jukka Lekkala,et al.  ElectroMechanical Film (EMFi) — a new multipurpose electret material , 2000 .

[2]  Mark S. Nixon,et al.  Comparing Different Template Features for Recognizing People by their Gait , 1998, BMVC.

[3]  M. D. Addlesee,et al.  The ORL Active Floor , 1997 .

[4]  Barbara Hammer,et al.  Relevance determination in Learning Vector Quantization , 2001, ESANN.

[5]  Robert Headon,et al.  Recognizing movements from the ground reaction force , 2001, PUI '01.

[6]  Dennis H. Rouvray,et al.  Beyond calculation: The next fifty years of computing , 1997 .

[7]  Yrjö Neuvo,et al.  FIR-median hybrid filters , 1987, IEEE Trans. Acoust. Speech Signal Process..

[8]  Jukka Riekki,et al.  FOOTSTEP IDENTIFICATION FROM PRESSURE SIGNALS USING HIDDEN MARKOV MODELS , 2003 .

[9]  Atsushi Sato,et al.  Generalized Learning Vector Quantization , 1995, NIPS.

[10]  Ron Kohavi,et al.  A Study of Cross-Validation and Bootstrap for Accuracy Estimation and Model Selection , 1995, IJCAI.

[11]  Thomas Villmann,et al.  Generalized relevance learning vector quantization , 2002, Neural Networks.

[12]  Gert Pfurtscheller,et al.  Automated feature selection with a distinction sensitive learning vector quantizer , 1996, Neurocomputing.

[13]  J. Riekki,et al.  Learning Vector Quantization in Footstep Identification , 2003 .

[14]  Gregory D. Abowd,et al.  The smart floor: a mechanism for natural user identification and tracking , 2000, CHI Extended Abstracts.

[15]  Philippe C. Cattin,et al.  Biometric authentication system using human gait , 2002 .

[16]  A. N. Rajagopalan,et al.  Gait-based recognition of humans using continuous HMMs , 2002, Proceedings of Fifth IEEE International Conference on Automatic Face Gesture Recognition.

[17]  Alan R. Jones,et al.  Fast Fourier Transform , 1970, SIGP.

[18]  Jukka Riekki,et al.  Reject-Optional LVQ-Based Two-Level Classifier to Improve Reliability in Footstep Identification , 2004, Pervasive.

[19]  Jorma Laaksonen,et al.  LVQ_PAK: The Learning Vector Quantization Program Package , 1996 .

[20]  J. Little,et al.  Recognizing People by Their Gait: The Shape of Motion , 1998 .

[21]  Jeha Ryu,et al.  The User Identification System Using Walking Pattern over the ubiFloor , 2003 .