A waveform concatenation technique for text-to-speech synthesis

Designing text-to-speech systems capable of producing natural sounding speech segments in different Indian languages is a challenging and ongoing problem. Due to the large number of possible pronunciations in different Indian languages, a number of speech segments are needed to be stored in the speech database while a concatenative speech synthesis technique is used to achieve highly natural speech segments. However, the large speech database size makes it unusable for small hand held devices or human computer interactive systems with limited storage resources. In this paper, we proposed a fraction-based waveform concatenation technique to produce intelligible speech segments from a small footprint speech database. The results of all the experiments performed shows the effectiveness of the proposed technique in producing intelligible speech segments in different Indian languages even with very less storage and computation overhead compared to the existing syllable-based technique.     

Modified neo-fuzzy neuron-based approach for economic and environmental optimal power dispatch

At the central energy management center in a power system, the real time controls continuously track the load changes and endeavor to match the total power demand with total generation in such a manner that the operating cost is minimized while all the operating constraints are satisfied. However, due to the strict government regulations on environmental protection, operation at minimum cost is no longer the only criterion for dispatching electrical power. The idea behind the environmentally constrained economic dispatch formulation is to estimate the optimal generation schedule of generating units in such a manner that fuel cost and harmful emission levels are both simultaneously minimized for a given load demand. Conventional optimization techniques become very time consuming and computationally extensive for such complex optimization tasks. These methods are hence not suitable for on-line use. Neural networks and fuzzy systems can be trained to generate accurate relations among variables in complex non-linear dynamical environment, as both are model-free estimators. The existing synergy between these two fields has been exploited in this paper for solving the economic and environmental dispatch problem on-line. A multi-output modified neo-fuzzy neuron (NFN), capable of real time training is proposed for economic and environmental power generation allocation.This model is found to achieve accurate results and the training is observed to be faster than other popular neural networks. The proposed method has been tested on medium-sized sample power systems with three and six generating units and found to be suitable for on-line combined environmental economic dispatch (CEED).     

Analyzing collective behavior from blogs using swarm intelligence

With the rapid growth of the availability and popularity of interpersonal and behavior-rich resources such as blogs and other social media avenues, emerging opportunities and challenges arise as people now can, and do, actively use computational intelligence to seek out and understand the opinions of others. The study of collective behavior of individuals has implications to business intelligence, predictive analytics, customer relationship management, and examining online collective action as manifested by various flash mobs, the Arab Spring (2011) and other such events. In this article, we introduce a nature-inspired theory to model collective behavior from the observed data on blogs using swarm intelligence, where the goal is to accurately model and predict the future behavior of a large population after observing their interactions during a training phase. Specifically, an ant colony optimization model is trained with behavioral trend from the blog data and is tested over real-world blogs. Promising results were obtained in trend prediction using ant colony based pheromone classier and CHI statistical measure. We provide empirical guidelines for selecting suitable parameters for the model, conclude with interesting observations, and envision future research directions.     

A comparative dielectric relaxation study of PMN–PT and PMN–PZ ceramics using impedance spectroscopy

AC-impedance spectroscopic studies in the temperature range of 30–400 °C are carried out on solid solutions of lead magnesium niobate (PMN) with lead titanate (PT) and lead zirconate (PZ), both of them in the 65/35 atomic ratio. For PMN–PT this corresponds to the morphotropic phase boundary composition (with normal ferroelectric behaviour), and for PMN–PZ it is near the phase boundary between normal ferroelectric and relaxor ferroelectric compositions. The variation of dielectric permittivity with temperature at different frequencies shows normal ferroelectric and relaxor-like dependence for PMN–PT and PMN–PZ, respectively. Temperature-dependent spectroscopic modulus plots reveal a much broader peak for PMN–PZ compared to that for PMN–PT, which is consistent with the dielectric behaviour. PMN–PT shows nearly ideal Debye behaviour below T_m (the temperature of the permittivity maximum) and the behaviour departs from ideality above T_m, whereas non-ideal Debye behaviour is seen both below and above T_m for PMN–PZ. Complex modulus plots fit well with two depressed semicircles and three depressed semicircles, respectively, for PMN–PT and PMN–PZ. The relaxation observed in the spectroscopic plots around 1 MHz for PMN–PT has been assigned to polarisation relaxation expected for normal-sized domains. No such relaxation could be observed for PMN–PZ around 1 MHz because of the mesoscopic domain sizes.     

A QoS-enhanced intelligent stochastic real-time packet scheduler for multimedia IP traffic

A re-configurable, QoS-enhanced intelligent stochastic real-time optimal fair packet scheduler, QUEST, for IP routers is proposed and investigated. The objective is to maximize the system QoS subject to the constraint that the processor utilization is kept at 100%. All past work on router schedulers for multimedia traffic were of earlier generation, in that they focused on maximizing utilization whereas being QoS-aware but without explicitly maximizing the QoS. Keeping utilization fixed at nearly 100%, QoS is dynamically maximized, thus moving to the next generation. QUEST’s other unique advantages are three-fold. First, it solves the challenging problem of starvation for low priority processes; second, it solves the major bottleneck of Earliest Deadline First scheduler’s failure at heavy traffic loads. Finally, QUEST offers the benefit of arbitrarily pre-programming the process utilization ratio. Three classes of multimedia IP traffic, namely, VoIP, IPTV and HTTP have been considered. Two most important QoS metrics, namely, packet loss rate (PLR) and mean waiting time, are addressed. All claims are supported by discrete event and Monte Carlo simulations. The proposed scheduler outperforms benchmark schedulers and offers 37% improvement in packet loss rate and 23% improvement in mean waiting time over the best competing current scheduler Accuracy-aware EDF. The proposed scheduler was validated in a test-bed platform of a NetFPGA® router and results were observed with Paessler® PRTG network monitor.

     

Cytochrome P450 inhibitory potential of selected Indian spices — possible food drug interaction

Spices constitute an important group of food which is virtually indispensable in the culinary art. In a view, these spices feared to pose a probability to affect the disposition of conventional pharmaceuticals through inhibition of human cytochrome P450 (CYPs) enzymes. In the present study an approach has been made to evaluate the possible CYP inhibition potential with some Indian spices (Capsicum annuum, Murraya koenigii, Zingiber officinale) and their major bioactive compounds, in combination with pooled microsome; as well as commercially available recombinant human CYP3A4, CYP2D6, CYP2C9 and CYP1A2. Quantification of the bioactive compound was determined through RP-HPLC, in order to standardize the plant material. CYP–carbon monoxide (CYP–CO) complex assay result indicated that all the plants and their bioactive compounds have an interaction potential with CYPs. Fluoregenic assay results indicated that the spice extracts have higher inhibition potential comparing to their single bioactive molecule. The higher enzyme inhibition potential by the extracts may be related to the synergistic effects due to the presence of other constituents in the extract. Capsaicin and C. annuum showed the lowest IC₅₀ value and 6-gingerol and Z. officinale extract showed the highest IC₅₀ value among the entire sample tested. The entire sample showed significantly less (P < 0.001, P < 0.01) interaction potential than known inhibitors. These findings indicate that selected spices are unlikely to cause clinically relevant drug interactions involving the inhibition of major CYP isozymes.     

Deep Scalogram Representations for Acoustic Scene Classification

Spectrogram representations of acoustic scenes have achieved competitive performance for acoustic scene classification. Yet, the spectrogram alone does not take into account a substantial amount of time-frequency information. In this study, we present an approach for exploring the benefits of deep scalogram representations, extracted in segments from an audio stream. The approach presented firstly transforms the segmented acoustic scenes into bump and morse scalograms, as well as spectrograms; secondly, the spectrograms or scalograms are sent into pre-trained convolutional neural networks; thirdly, the features extracted from a subsequent fully connected layer are fed into (bidirectional) gated recurrent neural networks, which are followed by a single highway layer and a softmax layer; finally, predictions from these three systems are fused by a margin sampling value strategy. We then evaluate the proposed approach using the acoustic scene classification data set of 2017 IEEE AASP Challenge on Detection and Classification of Acoustic Scenes and Events (DCASE). On the evaluation set, an accuracy of 64.0% from bidirectional gated recurrent neural networks is obtained when fusing the spectrogram and the bump scalogram, which is an improvement on the 61.0% baseline result provided by the DCASE 2017 organisers. This result shows that extracted bump scalograms are capable of improving the classification accuracy, when fusing with a spectrogram-based system.      

On the Role of C Addition on α Precipitation in a β Titanium Alloy

The onset of α nucleation in a carbon containing β-titanium alloy has been investigated by coupling atom probe tomography (APT) with transmission electron microscopy. The analysis of the APT results indicates that in addition to ω precipitates that can act as potential α nucleation sites, carbon atoms tend to form clusters within the β-matrix, which in turn give rise to additional nucleation sites for α, resulting in finer scale α precipitates due to increased nucleation density.