Journal 2008

INTERNATIONAL PROCEEDING

 MODELLING AIR POLLUTION PROBLEM BY CELLULARNEURAL NETWORK

Vu Duc Thai, Pham Thuong Cat

 

Abstract: Introduce the appy of CNN technology in simulating to solve air pullution problems range in distributed pollutant source.

Proceeding of ICARCV2008, Hanoi,Page.1115-1118 

REAL-TIME RECONSTRUCTION OF SYMMETRICAL IMAGE USING CELLULAR NEURAL NETWORK

Pham Duc Long, Pham Thuong Cat

Abstract: A real-time algorithm to reconstruct a 2D vertically symmetrical image using a Cellular Neural Network (CNN) is proposed in this paper. The processing speed of the proposed algorithm on a CNN machine is much faster in comparison with that of algorithms running serially on a traditional digital computer. The proposed algorithm serves as a basic routine in CNN algorithms’ Library or can be used to reconstruct damaged or partially known 2D symmetrical images in real-time applications. The algorithm is tested on some images with good results.

ICARCV 2008 HaNoi 3-2008, IEEE, pp.1128-1132. 

THE PERCEPTUAL WAVELET FEATURE FOR NOISE ROBUST VIETNAMESE SPEECH RECOGNITION

Phung Trung Nghia, Nguyen Quoc Trung

Abstract: This study proposed a novel noise robust speech feature for Vietnamese speech recognition based on the Bark scale, the Perceptual Wavelet Packet Transform (PWPT), the Wavelet sub-band Energy Parameters (WEP). This method was used in compare to the widely used MFCC feature recognizer. The experimental results show that our proposal is superior to the widely used MFCC feature for Vietnamese speech corpus with both clean and white noisy.

Proceeding of the IEEE International Conference on Communications and Electronics (HUT-ICCE 2008), Hoi An, Vietnam, pp. 258 – 261, 6/2008 

A NOVEL FAST NOISE ROBUST VIETNAMESE SPEECH RECOGNITION APPLIED FOR ROBOT CONTROL

Phung Trung Nghia, Thai Quang Vinh

Abstract: Most of researches on speech recognition in the world concentrate on improving the large vocabulary of the corpus. In real-time robot control by speech commands, speech recognition is usually no need very large vocabulary but the fast implementation and the noise robust is prerequisite. This study proposes a novel fast noise robust wavelet-based Vietnamese speech recognition applied for robot control.  The experimental results of a robot control simulations built in MATLAB show that our proposal is superior to the other conventional speech recognizers for Vietnamese speech corpus with clean and white noisy speech. In addition, the computation cost of our proposal is much lower than the others.

Proceeding of the IEEE International Conference on Control, Automation, Robotics and Computer Vision (ICARCV 2008), pp. 821-826, Ha Noi, Vietnam, 11/2008

 A NEW WAVELET-BASED WIDE BAND SPEECH CODER

Phung Trung Nghia, Do Dinh Cuong, Pham Viet Binh 

Abstract: There are several methods and standards for speech coding.  Most of them are used for narrowband speech.  In multimedia communication systems, wideband speech coding plays an important role. Wavelet is an efficient signal processing tool for speech coding.  Conventional wavelet speech coders use wavelet global or sub-band dependent threshold to allocate bits in each sub-band. Using psychoacoustic model with temporal and simultaneous properties, we will be able to estimate the threshold close to human hearing. This paper presents a wavelet packet based wideband speech coding incorporating both backward temporal, forward temporal  and simultaneous masking models. The coder uses also Huffman  lossless compression. By applying this model we were calculated  the  bit  rate  results  of  approximately  25  kbps  while  preserving perceptual quality with single channel wide-band speech sampled at 16 KHz.

Proceeding of the IEEE International Conference on Advanced Technologies for Communications (ATC08), pp. 349-352, Hanoi, Vietnam, 10/2008. 

Analysis and identification of Peer-to-Peer traffic in the WIDE backbone

Vu Thanh Vinh, Binh Pham Viet, Hung Nguyen Chan 

Abstract: Nowadays, P2P applications account for very high percentage of Internet traffic. As a result, P2P traffic identification is very important in network planning, security, QoS, etc. In particular, ISPs can utilize this data to adapt their policy in order to improve network performance. In this paper, we introduce a new method of P2P traffic identification base on flow timeout feature which can increase the accuracy of the identification process up to 100% in comparison with traditional methods. The experiments were performed base on trace data of WIDE (Widely Integrated Distributed Environment) backbone, the trans-Pacific backbone connect Japan and North America. Some characteristics of P2P traffic of today Internet are also analyzed and given in this study

Proceedings of the 3rd International Conference on Ubiquitous Information Technologies and Appliction (ICUT -3), Ho Chi Minh, Vietnam, 2008. Pp 294-299

Download Journal 2008