Fisher Discriminative Dictionary Learning for Vehicle Classification in Acoustic Sensor Networks

In this paper, we propose a discriminative dictionary learning framework for vehicle classification to improve the classification accuracy, and study the fisher discriminative dictionary learning (FDDL) approach in acoustic sensor networks. More precisely, the acoustic sensor data sets are captured to measure the vehicle running event. The multi-dimensional frequency spectrum features of sensor data sets are extracted using Mel frequency cepstral coefficients (MFCC), and the vehicle classification scheme is solved using fisher discriminative dictionary learning method, which exploits the discriminative information in both the representation residuals and the representation coefficients. To further analyze the performance of our proposed model, we extend our model to deal with sparse environmental noise. Extensive experiments are conducted on acoustic sensor databases and the results demonstrate that our proposed model shows superior performance in this vehicle classification framework compared to SVM, SRC, KSRC and LC-KSVD algorithms.