Browsing by Author "Chen, PC"

Now showing 1 - 9 of 9

A graded approach to shape representation
(1996-06-01)
Chen, PC
;
Tsai, WC
;
Hwang, SY
;
交大名義發表
;
資訊工程學系
Human beings perceived real-world objects in a graded manner, with macroviews providing global information about the objects and microviews providing more detailed information. Consequently, researchers in computer vision have directed their efforts toward building graded representations of objects, especially shapes of objects. Most previous approaches employed operations such as smoothing, blurring, or grouping to gradually obtain representations with different degrees of detail. Parameters of these operations are used to indicate the degree of detail, One objection to using operator parameters is that these parameters are difficult for viewers to interpret intuitively. To overcome this objection, we propose using a graded approach guided by a given similarity requirement to construct a more concise shape description. Our shape representation matches human intuition about detailing degree and suits machine requirement for concise description. (C) 1996 Academic Press, Inc.
A graded approach to shape representation
(1996-06-01)
Chen, PC
;
Tsai, WC
;
Hwang, SY
;
資訊工程學系
;
Department of Computer Science
Human beings perceived real-world objects in a graded manner, with macroviews providing global information about the objects and microviews providing more detailed information. Consequently, researchers in computer vision have directed their efforts toward building graded representations of objects, especially shapes of objects. Most previous approaches employed operations such as smoothing, blurring, or grouping to gradually obtain representations with different degrees of detail. Parameters of these operations are used to indicate the degree of detail, One objection to using operator parameters is that these parameters are difficult for viewers to interpret intuitively. To overcome this objection, we propose using a graded approach guided by a given similarity requirement to construct a more concise shape description. Our shape representation matches human intuition about detailing degree and suits machine requirement for concise description. (C) 1996 Academic Press, Inc.
Adaptive AR modeling in white Gaussian noise
(1997-05-01)
Wu, WR
;
Chen, PC
;
交大名義發表
;
電信工程研究所
;
National Chiao Tung University
Autoregresssive (AR) modeling is widely used in signal processing, The coefficients of an AR model can be easily obtained with a least mean square (LMS) prediction error filter, However, it is known that this filter gives a biased solution when the input signal is corrupted by white Gaussian noise, Treichler suggested the gamma-LMS algorithm to remedy this problem and proved that the mean weight vector can converge to the Wiener solution. In this paper, we develop a new algorithm that extends works of Vijayan et al, for adaptive AR modeling in the presence of white Gaussian noise, By theoretical analysis, we show that the performance of the new algorithm is superior to the gamma-LMS filter, Simulations are also provided to support our theoretical results.
Adaptive AR modeling in white Gaussian noise
(1997-05-01)
Wu, WR
;
Chen, PC
;
電信工程研究所
;
Institute of Communications Engineering
Autoregresssive (AR) modeling is widely used in signal processing, The coefficients of an AR model can be easily obtained with a least mean square (LMS) prediction error filter, However, it is known that this filter gives a biased solution when the input signal is corrupted by white Gaussian noise, Treichler suggested the gamma-LMS algorithm to remedy this problem and proved that the mean weight vector can converge to the Wiener solution. In this paper, we develop a new algorithm that extends works of Vijayan et al, for adaptive AR modeling in the presence of white Gaussian noise, By theoretical analysis, we show that the performance of the new algorithm is superior to the gamma-LMS filter, Simulations are also provided to support our theoretical results.
Frame-based subband Kalman filtering for speech enhancement
(1998)
Wu, WR
;
Chen, PC
;
Chang, HT
;
Kuo, CH
;
電信工程研究所
Kalman Filtering is an effective speech enhancement technique, in which speech and noise signals are usually modeled as autoregressive (AR) processes and represented in the state-space domain. Since AR coefficients identification and Kalman filtering require extensive computations, practical implementation of this approach is difficult. This paper proposes a simple and practical scheme that overcomes these problems. Speech signals are first decomposed into subbands. Subband speech signals are then modeled as low-order AR processes, such that low-order Kalman filters can be applied. Enhanced fullband speech signals are finally obtained by combining the enhanced subband speech signals. Using a frame-based algorithm, autocorrelation functions of subband speech are calculated and the Yuler-Walker equations are solved to obtain the AR parameters. Simulation results show that Kalman filtering in the subband domain not only greatly reduces the computational complexity but also achieves better performance compared to that in the fullband domain.
Nanoimprinting-lithography-induced self-aligned liquid crystals for novel multifunctional optical films
(2006-02-13)
Chiu, CH
;
Kuo, HL
;
Chen, PC
;
Wen, CH
;
Liu, YC
Polymerizable nematic liquid crystal (LC) was adopted as a resist for our newly developed nanoimprinting lithography process. Liquid crystal molecules were self-aligned during imprinting process without precoated alignment layer. The large patterned area was able to achieve 4x4 cm(2) at very low impriting pressure, 1.5 bar. The optical anisotropic was verified through polarized optical microscopy and the retardation was measured by polarized ultraviolet/visible spectrometer. The micrograted LC films prepared by this novel method offer the potential applications for a LC alignment layer within LC cells, as well as preparation of optical anisotropic films. (c) 2006 American Institute of Physics.
Schedulers for 1xEV-DO: Third generation wireless high-speed data systems
(2003)
Huang, CY
;
Su, HY
;
Vitebsky, S
;
Chen, PC
;
交大名義發表
Scheduling algorithms have been developed and proposed to improve system performance by taking the benefit of channel variance among users. When the optimization of the system throughput is the main objective for scheduler algorithm designs, the associated algorithms might cause unfairness in user transmission rates. To explore the fairness problem, in this paper, we will focus on the fairness performance by considering different scheduling algorithms in various fading channels. The Dynamic Target with Max/Min (DTM) Rate Controlled Scheduler Algorithm is proposed to resolve the potential fairness problems and to maintain the minimum transmission rate for all uses in which we believe that most applications require the minimum rate to maintain the application quality. The DTM algorithm includes also the maximum rate control to prevent the resource hogging from users at the good RF conditions.
Subband Kalman filtering for speech enhancement
(1998-08-01)
Wu, WR
;
Chen, PC
;
電信工程研究所
;
Institute of Communications Engineering
Kalman filtering is an effective speech-enhancement technique, in which speech signals are usually modeled as autoregressive (AR) processes and represented in the state-space domain. Since AR coefficients identification and Kalman filtering require extensive computations, real-time implementation of this approach is difficult. This paper proposes a simple and practical scheme that overcomes these obstacles. Speech signals are first decomposed into subbands. Subband speech signals are then modeled as low-order AR processes, such that low-order Kalman filters can he applied. Enhanced fullband speech signals are finally obtained by combining the enhanced subband speech signals. To identify AR coefficients, prediction-error filters adapted by the LMS algorithm are applied. Due to noisy inputs, the LMS algorithm converges to biased solutions. The performance of the Kalman filter with biased parameters is analyzed, It is shown that accurate estimates of AR coefficients are not required when the driving-noise variance is properly estimated. New methods for making such estimates are proposed. Thus, we can tolerate biased AR coefficients and take advantage of the LMS algorithm's simple structure. Simulation results show that speech enhancement in the subband domain not only greatly reduces the computational complexity, but also achieves better performance compared to that in the fullband domain.
The effect of barrier structure on the performance of double barrier quantum well infra-red photodetectors
(1996-02-01)
Tsai, KL
;
Lee, CP
;
Chen, PC
;
Tsang, JS
;
Tsai, CM
The effects of the inner and outer barriers on the performance of mid-wavelength (3-5 mu m) GaAs/AlGaAs double barrier quantum well infra-red photodetectors (DBQWIPs) have been studied. It was found that the peak response wavelength had a significant red shift when the thickness of the AlAs inner barriers was below 10 Angstrom. This is attributed to the lowering of the upper state energy in the quantum wells. The responsivity of DBQWIPs decreases with the inner barrier thickness. This is because the optical gain decreases faster than the increasing rate of the quantum efficiency when the inner barrier thickness increases. The responsivities for devices with different Al compositions (0.35 and 0.24) in the outer barriers are in the same range, but the dark currents for devices with Al(0.35)G(0.65)As outer barriers are much lower than those with Al0.24Ga0.76As outer barriers. Nearly two orders of magnitude improvement in detectivity was observed when the Al composition in the outer barriers was changed from 0.24 to 0.35. A high detectivity of 1 x 10(11) cm Hz(1/2) W-1 at 77 K was obtained for a device with 12 Angstrom AlAs inner barriers and 300 Angstrom Al0.35Ga0.65As outer barriers.