Deep Cellular Recurrent Network for Efficient Analysis of Time-Series Data With Spatial Information.

Efficient processing of large-scale time-series data is an intricate problem in machine learning. Conventional sensor signal processing pipelines with hand-engineered feature extraction often involve huge computational costs with high dimensional data. Deep recurrent neural networks have shown promise in automated feature learning for improved time-series processing. However, generic deep recurrent models grow in scale and depth with the increased complexity of the data. This is particularly challenging in presence of high dimensional data with temporal and spatial characteristics. Consequently, this work proposes a novel deep cellular recurrent neural network (DCRNN) architecture to efficiently process complex multidimensional time-series data with spatial information. The cellular recurrent architecture in the proposed model allows for location-aware synchronous processing of time-series data from spatially distributed sensor signal sources. Extensive trainable parameter sharing due to cellularity in the proposed architecture ensures efficiency in the use of recurrent processing units with high-dimensional inputs. This study also investigates the versatility of the proposed DCRNN model for the classification of multiclass time-series data from different application domains. Consequently, the proposed DCRNN architecture is evaluated using two time-series data sets: a multichannel scalp electroencephalogram (EEG) data set for seizure detection, and a machine fault detection data set obtained in-house. The results suggest that the proposed architecture achieves state-of-the-art performance while utilizing substantially less trainable parameters when compared to comparable methods in the literature.

Sparse Simultaneous Recurrent Deep Learning for Robust Facial Expression Recognition.

Facial expression recognition is a challenging task that involves detection and interpretation of complex and subtle changes in facial muscles. Recent advances in feed-forward deep neural networks (DNNs) have offered improved object recognition performance. Sparse feature learning in feed-forward DNN models offers further improvement in performance when compared to the earlier handcrafted techniques. However, the depth of the feed-forward DNNs and the computational complexity of the models increase proportional to the challenges posed by the facial expression recognition problem. The feed-forward DNN architectures do not exploit another important learning paradigm, known as recurrency, which is ubiquitous in the human visual system. Consequently, this paper proposes a novel biologically relevant sparse-deep simultaneous recurrent network (S-DSRN) for robust facial expression recognition. The feature sparsity is obtained by adopting dropout learning in the proposed DSRN as opposed to usual handcrafting of additional penalty terms for the sparse representation of data. Theoretical analysis of S-DSRN shows that the dropout learning offers desirable properties such as sparsity, and prevents the model from overfitting. Experimental results also suggest that the proposed method yields better performance accuracy, requires reduced number of parameters, and offers reduced computational complexity than that of the previously reported state-of-the-art feed-forward DNNs using two of the most widely used publicly available facial expression data sets. Furthermore, we show that by combining the proposed neural architecture with a state-of-the-art metric learning technique significantly improves the overall recognition performance. Finally, a graphical processing unit (GPU)-based implementation of S-DSRN is obtained for real-time applications.

Real-Time Epileptic Seizure Detection Using EEG.

This paper proposes a novel patient-specific real-time automatic epileptic seizure onset detection, using both scalp and intracranial electroencephalogram (EEG). The proposed technique obtains harmonic multiresolution and self-similarity-based fractal features from EEG for robust seizure onset detection. A fast wavelet decomposition method, known as harmonic wavelet packet transform (HWPT), is computed based on Fourier transform to achieve higher frequency resolutions without recursive calculations. Similarly, fractal dimension (FD) estimates are obtained to capture self-similar repetitive patterns in the EEG signal. Both FD and HWPT energy features across all EEG channels at each epoch are organized following the spatial information due to electrode placement on the skull. The final feature vector combines feature configurations of each epoch within the specified moving window to reflect the temporal information of EEG. Finally, relevance vector machine is used to classify the feature vectors due to its efficiency in classifying sparse, yet high-dimensional data sets. The algorithm is evaluated using two publicly available long-term scalp EEG (data set A) and short-term intracranial and scalp EEG (data set B) databases. The proposed algorithm is effective in seizure onset detection with 96% sensitivity, 0.1 per hour median false detection rate, and 1.89 s average detection latency, respectively. Results obtained from analyzing the short-term data offer 99.8% classification accuracy. These results demonstrate that the proposed method is effective with both short- and long-term EEG signal analyzes recorded with either scalp or intracranial modes, respectively. Finally, the use of less computationally intensive feature extraction techniques enables faster seizure onset detection when compared with similar techniques in the literature, indicating potential usage in real-time applications.