One of the merits of Free-Viewpoint TV (FTV) is to generate a scene according to the user's viewpoint. In order to generate a scene according to the user's requirements, a position in 3D space has to be defined as the viewpoint. To achieve this, we present an intelligent remote controller that locate a focus of attention for the generation of candidate viewpoints based on the theory of human visual attention. It enables a viewer to change his viewpoints with easy by providing some candidate viewpoints automatically, and also prevents the viewer from disorientation caused occasionally from 3D space navigation.
This paper presents a novel token-based approach for spoken language identification (LID) using bayesian logistic regression model, which takes into account prior distribution for parameters of logistic regression models in order to avoid overfitting. Speech utterances are first decoded into token sequences, and then we design a hierarchical system which utilizes bayesian logistic regression model to perform LID task on these token sequences. Experiments conducted on the NIST LRE 2007 database show that the proposed approach provides quite competitive performance compared to other state-of-the-art token-based approaches.
The first step in membrane protein type prediction is how to represent the sample of a protein. In this work, a protein can be represented by a high-dimension feature vector by using the DC (Dipeptide Composition) method. This extremely high dimensionality of the protein data may increase the computing time and classifier complexity. Thus, a linear dimensionality reduction algorithm NPE (Neighborhood Preserving Embedding) is introduced to extract the indispensable features from the high-dimensional DC space. Based on the reduced low-dimensional features, K-NN (K-nearest neighbor) classifier is introduced to identify the types of membrane proteins. Finally, a very encouraging experimental result is obtained.
A novel discrete dimmable electronic ballast for a fluorescent lamp without changing the original power wiring system is developed. We propose a new method that detects the number of times the power-switch is pressed to make the electronic ballast generate the corresponding power rating. A dimming module realizes the proposed method and makes the DC-AC inverter in a round-robin manner to output the respective power rating. A prototype of the discrete dimmable electronic ballast for a T5 fluorescent lamp of 28W under 110VAC/60Hz, for generating a four-state power rating is developed to demonstrate the feasibility of this approach.
Several field experiments were conducted to detect and track ships using a multi-frequency high-frequency surface-wave radar (HFSWR) system on the coast of East China Sea during 2009. Shipborne Automatic Identification System (AIS) was used to verify the results. And the algorithm of detection and tracking was proved to be efficient and practical.
This paper presents a clock frequency doubler, having the function of automatic adjustable duty cycle without feedback loops. The duty cycle amount can be automatically adjustable using digitized delay block and a counter. This simplifies the design structure and allows the circuit to operate over a wide range of input frequency variation. The simulation results show that this frequency doubler operates at a very wide variable input frequency ranging from 650MHz to 1.25GHz.
New signal modulation design for next-generation of global navigation satellite systems (GNSS) must offer improved performance and the opportunity for spectrum compatible with existing and planned signals. This paper proposes a class of particularly attractive modulations called Minimum Shift Keying - Binary Coded Symbol (MSK-BCS) modulations, and demonstrates their essential characteristics and performance for GNSS. The principle of MSK modulation is first described. Then, based on the BCS modulations technique, the general mathematical model of MSK-BCS modulations is described. The MSK-BCS modulations present important characteristics for GNSS and provide superior performance with regards to noise, interference and multipath when compared to other BCS modulations. MSK-BCS modulations can provide potential opportunities for GNSS modernization and construction, such as Galileo, Compass, and so on.
This paper presents the first successful demonstration of all-optical MEMS (microelectromechanical system) fiber endoscope for OCT (optical coherence tomography) measurement. A MEMS optical scanner for spatial light modulation has been assembled in the endoscope probe head (φ6mm) with optical components and a photovoltaic cell. Actuation energy for the MEMS scanner is optically supplied by using a 1.5µm light of 10mW, and OCT probing is operated by another wavelength at 1.3µm. Use of optical fiber for both powering and measurement makes a safe endoscope that is free from the risk of electrocution or electromagnetic interference.
A compact microstrip lowpass filter (LPF) with prominent cutoff sharpness and wide stopband is presented. The microstrip LPF is composed of a broadened microstrip line on the top side of the microstrip structure, an H-shaped defected ground structure (DGS) and a modified dumbbell (DB) DGS in the ground plane. It has a simple design procedure and easily can be tuned. The designed filter is developed and experimental results show that it has a sharp cutoff frequency response.
The OpenVG standard has been introduced as an efficient vector graphics API for embedded systems. There have been several OpenVG implementations that are based on the software rendering of image. However, the software rendering needs more execution time and power consumption than hardware accelerated rendering. For the efficient hardware implementation, we merge eight pipeline stages in the original specification to four pipeline stages. The first hardware acceleration stage is the tessellation part which is one of the pipeline stages that calculates the edge of vector graphics. In this paper, we provide an efficient hardware design for the tessellation stage and claim this would eventually reduce the execution time and hardware complexity.
In video coding, a portion of skipped macroblocks is highly effective to the coding performance, especially, at low bitrates since, for a skip mode, no additional coding parameters are transmitted except mode index itself. To increase the skip mode occurrence, in this paper, we propose an efficient skip mode competition scheme based on vector clustering of multiple predictive motion vectors (PMV) and boundary detection of moving objects. The proposed scheme selects the best PMV of skip mode optimally among representative PMV candidates which are obtained after the clustering process and switches a more appropriate competition scheme adaptively according to object boundary conditions. Experimental results show that the proposed scheme reduces average 4.31% in BDRATE compared to the KTA 2.6 software.
Instead of a bulky TE-mode resonator with high Qu, a compact TM-mode resonator with moderate Qu is proposed. The contact between the ceramic rod and the metal cavity is a key point for the filter reliability. Elasticity of the contact portion is well utilized in this case to keep the smooth electrical contact. Also, temperature coefficient of TM-mode resonator is studied. The requirements for characteristics of the ceramic materials are a little different for TE-mode and TM-mode. Finally, high-reliable and temperature-stable TM-mode resonator has been developed. Its application to compact base station filter is demonstrated as well.
Calibration of an electric field probe using a different type and size of probe as a reference is carried out in a GTEM cell from 1 to 6GHz. We compared our results with those of calibration in an anechoic chamber to determine the effectiveness of the calibration. The difference between the calibration factors of the GTEM cell and the anechoic chamber increases when the probe is calibrated in an area in which the electric field distribution is nonuniform. Therefore, it is important to place the probe in an area with good uniformity or maintain the uniformity of the electric field in the GTEM cell when calibrating electric field probes with different dimensions.
Safety measures concerning drunk driving are the major technical challenges in the car industry. We have developed systems to prevent drunk driving by applying our new, non-contact breath-alcohol sensor in which our original breath sensor is coupled with an alcohol sensor. The breath sensor detects breath by measuring electric currents of positively or negatively charged water clusters in breath that are separated by using an electric field. Our breath-alcohol sensor simultaneously detects the electrical signal of breath and alcohol in breath. This insures that the sample is of a person's breath, not an artificial source.