Sign Language Gesture Recognition with Bispectrum Features Using SVM

EasyChair Preprint 2166

Abstract

Wi-Fi based sensing system captures the signal reflections due to human gestures as Channel State Information (CSI) values in subcarrier level for accurately predicting the fine-grained gestures. The proposed work derives Bispectrum Features (BF) from raw signal values by adopting a Conditional Informative Feature Extraction (CIFE) technique derived from information theory to form a subset of informative and best features. Support Vector Machine (SVM) classifier is adopted in the present work to for classifying the gesture and to measure the prediction accuracy. The present work is validated on a secondary dataset, SignFi, having data collected from different environments with varying number of users and sign gestures. SVM reports an overall accuracy of 83.84%, 94.13%, 74.85% and 75.56% in different environments/scenarios.

Keyphrases: CSI, SVM, Wi-Fi, bispectrum, gesture recognition

@booklet{EasyChair:2166,
  author    = {Hasmath Farhana Thariq Ahmed and Hafisoh Ahmad and Phang Swee King and C V Aravind and Harkat Houda and Kulasekharan Narasingamurthi},
  title     = {Sign Language Gesture Recognition with Bispectrum Features Using SVM},
  howpublished = {EasyChair Preprint 2166},
  year      = {EasyChair, 2019}}