System and method for user recognition using motion sensor data

A system, comprising:

• a network communication interface;
• a computer-readable storage medium;
• a processor configured to interact with the network communication interface and the computer readable storage medium and execute one or more software modules stored on the storage medium, comprising:
• a software module that when executed configures the processor to receive an access request from a mobile device and to obtain a motion signal associated with a user captured by at least one motion sensor communicatively coupled to the mobile device;
• a feature extraction module that when executed configures the processor to:
• extract sets of features from the captured motion signal using a plurality of feature extraction algorithms, wherein each individual set among the sets of extracted features includes features extracted from the captured motion signal by a respective feature extraction algorithm of the feature extraction module, and wherein the plurality of feature extraction algorithms comprise at least one of Mel Frequency Cepstral Coefficients (MFCC), Shifted Delta Cepstral (SDC), Histogram of Oriented Gradients (HOG), Markov Transition Matrix, and deep embeddings extracted with Convolutional Neural Networks (CNN);
• and
• a feature selection module that when executed configures the processor to select a subset of discriminative features from the sets of extracted features, wherein the feature selection module comprises a principal component analysis algorithm;
• and
• a classification module that when executed configures the processor to classify a user as a genuine user or an imposter user based on a classification score generated by one or more classifiers of the classification module from an analysis of the subset of discriminative features, wherein the classification module when executed configures the processor to classify the subset of discriminative features using a stacked generalization technique, and wherein the stacked generalization technique utilizes one or more of the following classifiers:
• (1) Naïve Bayes classifier, (2) Support Vector Machine (SVM) classifier, (3) Multi-layer Perception classifier, (4) Random Forest classifier, (5) and Kernel Ridge Regression (KRR), wherein
• the software module, when executed, further configures the processor to transmit an access grant to the mobile device in response to classifying the user as the genuine user and to transmit an access denial to the mobile device in response to classifying the user as the imposter user.