In this paper, we address the speech-based gender identification problem. Mel-Frequency Cepstral Coefficients (MFCC) of voice samples are typically used as the features for gender identification. However, MFCC-based classification incurs high complexity. This paper proposes a novel pitch-based gender identification system with a two-stage classifier to ensure accurate identification and low complexity. The first stage of the classifier identifies and labels all the speakers whose pitch clearly indicates the gender of the speaker; the complexity of this stage is very low since only threshold-based decision rule on a scalar (i.e., pitch) is used. The ambiguous voice samples from all the other speakers (which cannot be classified with high accuracy by the first stage, and can be regarded as suspicious speakers or difficult cases) are forwarded to the second-stage for finer examination; the second-stage of our classifier uses Gaussian Mixture Model (GMM) to accurately isolate voice samples based on gender. Experiment results show that our system is speech language/content independent, microphone independent, and robust against noisy recording conditions. Our system is extremely accurate with probability of correct classification of 98.65%, and very efficient with about 5 seconds required for feature extraction and classification.

Genome-wide association studies (GWAS) on sporadic Parkinson's disease (sPD) are mainly conducted in European and American populations at present, and the Han populations of Chinese mainland (HPCM) almost have not been studied yet. Here, we conducted a pooling GWAS combining a pathway analysis with 862,198 autosomal single nucleotide polymorphisms of IlluminaHumanOmniZhongHua-8 in 250 sPD and 250 controls from HPCM precluded toxicant exposure, age, and heavy coffee drinking habit interference. We revealed that among the 22 potential loci implicated, PRDM2/KIAA1026 (kgp8090149), TSG1/MANEA (kgp154172), PDE10A (kgp8130520), MDGA2 (rs9323124), ATPBD4/LOC100288892 (kgp11333367), ZFP64/TSHZ2 (kgp4156164), PAQR3/ARD1B (kgp9482779), FLJ23172/FNDC3B (kgp760898), C18orf1 (kgp348599), FLJ43860/NCRNA00051 (kgp4105983), CYP1B1/C2orf58 (kgp11353523), WNT9A/LOC728728 (rs849898), ANXA1/LOC100130911 (rs10746953), FLJ35379/LOC100132423 (kgp9550589), PLEKHN1 (kgp7172368), DMRT2/SMARCA2 (kgp10769919), ZNF396/INO80C (rs1362858), C3orf67/LOC339902 (rs6783485), LOC285194/IGSF11 (rs1879553), FGF10/MRPS30 (rs13153459), BARX1/PTPDC1 (kgp6542803), and COL5 A2 (rs11186), the peak significance was at the kgp4105983 of FLJ43860 gene in chromosome 8, the first top strongest associated locus with sPD was PRDM2 (kgp8090149) in chromosome 1, and the 24 pathways including 100 significantly associated genes were strongly associated with sPD from HPCM. The 40 genes were shared by at least two pathways. The most possible associated pathways with sPD were axon guidance, ECM-receptor interaction, neuroactive ligand-receptor interaction, tight junction, focal adhesion, gap junction, long-term depression, drug metabolism-cytochrome P450, adherens junction, endocytosis, and protein digestion and absorption. Our results indicated that these loci, pathways, and their related genes might be involved in the pathogenesis of sPD from HPCM and provided some novel evidences for further searching the genetic pathogenesis of sPD.

The aim of this study was to investigate the effects of melatonin on oxidative stress, the expression of transient receptor potential melastatin-2 (TRPM2) in guinea pig brains, and the influence of melatonin on oxidative stress in lungs and airway inflammation induced by particulate matter 2.5 (PM2.5). A particle suspension (0.1 g/ml) was nasally administered to the guinea pigs to prepare a PM2.5 exposure model. Cough frequency and cough incubation period were determined through RM6240B biological signal collection and disposal system. Oxidative stress markers, including malondialdehyde (MDA), total antioxidant capacity (T-AOC), total superoxide dismutase (T-SOD), and glutathione peroxidase (GSH-Px), in the medulla oblongata were examined through spectrophotometer. Reactive oxygen species (ROS) were detected in the hypoglossal nucleus, cuneate nucleus, Botzinger complex, dorsal vagal complex, and airway through dihydroethidium fluorescence. Hematoxylin-eosin (HE) staining and substance P expression via immunohistochemistry revealed the inflammatory levels in the airway. TRPM2 was observed in the medulla oblongata through immunofluorescence and Western blot. The ultrastructure of the blood-brain barrier and neuronal mitochondria was determined by using a transmission electron microscope. Our study suggests that melatonin treatment decreased PM2.5-induced oxidative stress level in the brains and lungs and relieved airway inflammation and chronic cough. TRPM2 might participate in oxidative stress in the cough center by regulating cough.

Context-Aware Recommender System (CARS) aims to not only recommend services similar to those already rated with the highest score, but also provide opportunities for exploring the important role of temporal, spatial and social contexts for personalized web services recommendation. A key step for temporal-based CARS methods is to explore the time decay process of past invocation records to make the Quality of Services (QoS) prediction. However, it is a nontrivial task to model the temporal effects on web services recommendation, due to the dynamic features of contextual information in view of temporal spatial correlations. For instance, in locationaware services recommendation, the user's geographical position would change very frequently as time goes on. In this paper, we propose a Context-Aware Services Recommendation based on Temporal Effectiveness (CASR-TE) method. Inspired by existing time decay approaches, we first present an enhanced temporal decay model combining the time decay function with traditional similarity measurement methods. Then, we model temporal spatial correlations as well as their impacts on the user preference expansion. Finally, we evaluate the CASR-TE method on WS-Dream dataset by evaluation matrices of both RMSE and MAE. Experimental results show that our approach outperforms several benchmark methods with a significant margin.

Abstract-In this paper, we address the problem of large population speaker identification under noisy conditions. Major techniques for speaker identification is based on Mel-Frequency Cepstral Coefficients (MFCC), Gaussian Mixture Model (GMM) and Universal Background Model (UBM) which we call MFCC+GMM and MFCC+GMM+UBM. The approaches are known to perform very well for small population identification under low-noise conditions. However, the increase of population size can cause performance degradation of these schemes under noisy conditions. To mitigate this limitation, we propose a fuzzyclustering-based decision tree approach. The key idea of our approach is to 1) use a decision tree to hierarchically partition the whole population into groups of small size, and determine which speaker group at the leaf node a speaker under test belongs to, and 2) apply MFCC+GMM to the selected speaker group for speaker identification. The advantage of our approach is that we use features that are independent from MFCC to partition speakers into groups and only apply MFCC+GMM to speaker groups at the leaf level. The key challenge in our design is how to achieve a low error probability of decision-treebased classification. To address this, we adopt fuzzy clustering in constructing the tree for population partitioning, i.e., at each level, a speaker may belong to multiple groups. Such redundancy increases the probability of classifying a speaker under test into a correct group/node on the tree. Another novelty of this paper is that we use pitch and five vocal source features to construct a six-level decision tree. Experimental results demonstrate that our approach outperforms MFCC+GMM and MFCC+GMM+UBM with higher accuracy and lower complexity for large population identification under additive white Gaussian noise (AWGN) conditions.