Unsupervised speaker segmentation with residual phase and MFCC features

This paper proposes an unsupervised method for improving the automatic speaker segmentation performance by combining the evidence from residual phase (RP) and mel frequency cepstral coefficients (MFCC). This method demonstrates the complementary nature of speaker specific information present in the residual phase in comparison with the information present in the conventional MFCC. Moreover this method presents an unsupervised speaker segmentation algorithm based on support vector machine (SVM). The experiments show that the combination of residual phase and MFCC helps to identify more robustly the transitions among speakers.     

Calculation of the band structure and superconductivity in the hexagonal close-packed phase of silicon

We report a calculation of the band structure and superconductivity of silicon in the hexagonal close-packed phase under pressure. The effect of pressure on the band structure is obtained by means of the linear muffin-tin orbital method. The superconducting transition temperatureT _c is calculated using the formulas of Allen and Dynes¹ as well as those of McMillan². It is found that the value ofT _c increases with pressure up to 74.3 GPa. The increase inT _c is attributed to the continuous s d electron transfer under pressure. The calculated values ofT _c are compared with the available experimental data. Further, Heine's fifth power law³ has been tested from thed-band widths obtained from the band-structure results.     

Macro- and Microstructural Studies of Laser-Processed WE43 (Mg-Y-Nd) Magnesium Alloy

The macro- and microstructural changes in the fusion zone (FZ) of WE43 magnesium alloy processed by a diode-pumped ytterbium (IPG YLS-3000) continuous wave fiber laser for specified processing conditions (from 4.17 to 12.5 × 10⁷ J/m²) were studied. With the aid of computational heat-transfer model, the temperature history and cooling rate were determined for different laser-processing conditions. The computational heat-transfer results of laser-processed samples were used to correlate with microstructures characterized using optical, scanning, and transmission electron microscopies, and electron backscatter and X-ray diffraction analyses. The microhardness measurement was carried out to establish the structure–property relationship, and the results indicated that the minimal hardness variation (1 pct) within laser FZ was due to a constant heat extraction time (0.1 second), narrow variation in grain size (4 to 7 μm), and the type of precipitate (β-phase) formation.     

Fovea intensity comparison code for person identification and verification

In this paper a method is proposed for person identification and verification. The method proposed in Viola and Jones (2001) is used to detect the face region in the image. The detected face region is processed to determine the locations of the eyes and mouth. The facial and mouth features are extracted relative to the locations of the eyes and mouth. A new feature called fovea intensity comparison code (FICC) is obtained from intensity values of the face/mouth region. The dimension of the FICC is reduced using principal component analysis (PCA). Euclidean distance matching is used for identification and verification. The performance of the system is evaluated in real time in the laboratory environment, and the system achieves a recognition rate (RR) of 99.0% and an equal error rate (EER) of about 0.84% for 50 subjects. The performance of the system is also evaluated for the eXtended Multi Modal Verification for Teleservices and Security (XM2VTS) database, and the system achieves a recognition rate of 100% an equal error rate (EER) of about 0.23%.     

Concentration of n‐3 polyunsaturated fatty acids in cholesterol‐reduced cod‐liver oil by lipases

This study was conducted to elucidate the effects of different lipases originated from Candida rugosa (CR), porcine pancreas (PP) and Aspergillus niger (AN) on the degree of hydrolysis (DH) in cholesterol‐reduced cod‐liver oil (87.5%) and evaluate the changes in n‐3 polyunsaturated fatty acid concentrations in the oil hydrolysed by the lipases. The lipase‐catalysed hydrolysis of cholesterol‐reduced cod‐liver oil was performed at 37 °C for 8 h. Among all the lipase samples studied, DH in the oil after lipase‐catalysed hydrolysis followed the decreasing order: CR lipase (70.01%) > PP lipase (26.18%) > AN lipase (18.57%). Triacylglycerol levels in the oil hydrolysed by all the lipases studied decreased, while mono‐ and diacylglycerol levels increased during lipase‐catalysed hydrolysis. The eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) concentrations in cholesterol‐reduced cod‐liver oil hydrolysed by the CR lipase were remarkably higher than those by the PP or AN lipase. Thus, it is suggested in this study that the CR lipase appears to be most suitable for producing and n‐3 polyunsaturated fatty acids including EPA and DHA concentrates from cholesterol‐reduced cod‐liver oil.     

Physicochemical and antioxidative properties of Korean nanopowdered white ginseng

This study was carried out to investigate physicochemical and antioxidative properties of Korean nano white ginseng powder. The particle size of nanopowdered ginseng (NPG) was in the range from 600 to 1000 nm which was significantly smaller than the regular powdered ginseng (PG) which was in the range from 300 to 500 μm. Carbohydrate content was found to be significantly higher in NPG than PG (P < 0.05). Higher L value of NPG was mostly due to the higher light‐scattering effect of bright ginseng powder. In oil‐holding capacity, NPG was significantly higher than PG (P < 0.05). Total polyphenol content was not significantly different between NPG and PG, and it was in the range of 52.5% (P > 0.05). However, DPPH and ABTS studies showed that NPG has higher antioxidative properties than PG. In overall, NPG showed much higher antioxidative properties compared with PG. Thus, NPG enhances the functional value of Korean white ginseng powder.