High quality voice conversion using prosodic and high-resolution spectral features

Voice conversion methods have advanced rapidly over the last decade. Studies have shown that speaker characteristics are captured by spectral feature as well as various prosodic features. Most existing conversion methods focus on the spectral feature as it directly represents the timbre characteristics, while some conversion methods have focused only on the prosodic feature represented by the fundamental frequency. In this paper, a comprehensive framework using deep neural networks to convert both timbre and prosodic features is proposed. The timbre feature is represented by a high-resolution spectral feature. The prosodic features include F0, intensity and duration. It is well known that DNN is useful as a tool to model high-dimensional features. In this work, we show that DNN initialized by our proposed autoencoder pretraining yields good quality DNN conversion models. This pretraining is tailor-made for voice conversion and leverages on autoencoder to capture the generic spectral shape of source speech. Additionally, our framework uses segmental DNN models to capture the evolution of the prosodic features over time. To reconstruct the converted speech, the spectral feature produced by the DNN model is combined with the three prosodic features produced by the DNN segmental models. Our experimental results show that the application of both prosodic and high-resolution spectral features leads to quality converted speech as measured by objective evaluation and subjective listening tests.     

Enhanced security in internet voting protocol using blind signature and dynamic ballots

In this paper, we introduce an internet voting protocol which satisfies desired security requirements of electronic voting. In the newly proposed protocol, we allow the adversaries to get more power than in any previous works. They can be coercers or vote buyers outside, and corrupted parties inside our system. These adversaries also have ability to collude with each other to ruin the whole system. Our main contribution is to design an internet voting protocol which is unsusceptible to most of sophisticated attacks. We employ the blind signature technique and the dynamic ballots instead of complex cryptographic techniques to preserve privacy in electronic voting. Moreover, we also aim at the practical system by improving the blind signature scheme and removing physical assumptions which have often been used in the previous works.     

The fatigue life of contoured cobalt chrome posterior spinal fusion rods

Intraoperative contouring of posterior rods in lumbar arthrodesis constructs introduces stress concentrations that can substantially reduce fatigue life. The sensitivity of titanium (Ti) and stainless steel (SS) to intraoperative contouring has been established in the literature; however, notch sensitivity has yet to be quantified for cobalt chrome (CoCr), which is now being advocated for use in posterior arthrodesis constructs. The goal of this study is to evaluate the sensitivity of CoCr rods to intraoperative contouring for posterior lumbar screwrod arthrodesis constructs. In this paper lumbar bilateral vertebrectomy models are constructed based on ASTM F1717-01 with curved rods (26-30 degrees total curvature) and poly-axial pedicle screws. Three types of constructs are assembled: first, 5.5 mm SS rods with SS screws (6.5 x 35 mm), second, 6.0 mm Ti rods with Ti screws (7.5 x 35 mm), and third, 6.0 mm CoCr rods with Ti screws (7.5 x 35 mm). All specimens are tested at 4 Hz in dynamic axial compression-bending with a load ratio of ten and maximum load levels of 250, 400, and 700 N until run-out at 2 000 000 cycles. Results are presented that show that the fatigue life of CoCr constructs tend to be greater than Ti constructs at all levels. At the 400 N maximum loading, CoCr lasts an average of 350 000 cycles longer than the Ti constructs. The CoCr constructs are able to sustain the 250 N load until run-out at 2 000 000 cycles but they fail at high load levels (maximum 700 N). The CoCr constructs fail at the neck of the Ti screw at high loads whereas Ti screws fail at the notch induced by contouring. Since CoCr is compatible with magnetic resonance imaging and has high static strength characteristics, the results of this study suggest that it may be an appropriate substitute for Ti.     

Flow measurements in sewers based on image analysis: automatic flow velocity algorithm

Discharges of combined sewer overflows (CSOs) and stormwater are recognized as an important source of environmental contamination. However, the harsh sewer environment and particular hydraulic conditions during rain events reduce the reliability of traditional flow measurement probes. An in situ system for sewer water flow monitoring based on video images was evaluated. Algorithms to determine water velocities were developed based on image-processing techniques. The image-based water velocity algorithm identifies surface features and measures their positions with respect to real world coordinates. A web-based user interface and a three-tier system architecture enable remote configuration of the cameras and the image-processing algorithms in order to calculate automatically flow velocity on-line. Results of investigations conducted in a CSO are presented. The system was found to measure reliably water velocities, thereby providing the means to understand particular hydraulic behaviors.     

Topology optimization of a cantilevered piezoelectric energy harvester using stress norm constraints

Vibrational piezoelectric energy harvesters are devices which convert ambient vibrational energy into electric energy. Here we focus on the common cantilever type in which an elastic beam is sandwiched between two piezoelectric plates. In order to maximize the electric power for a given sinusoidal vibrational excitation, we perform topology optimization of the elastic beam and tip mass by means of the SIMP approach, leaving the piezoelectric plates solid. We are interested in the first and especially second resonance mode. Homogenizing the piezoelectric strain distribution is a common indirect approach increasing the electric performance. The large design space of the topology optimization approach and the linear physical model also allows the maximization of electric performance by maximizing peak bending, resulting in practically infeasible designs. To avoid such problems, we formulate dynamic piezoelectric stress constraints. The obtained result is based on a mechanism which differs significantly from the common designs reported in literature.     

VICUR: A human-vision-based algorithm for curve reconstruction

In this paper, a new algorithm, named VICUR, is presented for curve reconstruction problem. From a set of unorganized points, the proposed algorithm can construct curves that look natural to human vision. The VICUR algorithm is based on two connectivity criteria: proximity and good continuation from the prominent Gestalt principles of perception. Experimental results are presented to show the effectiveness of VICUR.