Toeplitz Robust Noisy Speech Endpoint Detection

2012 ◽  
Vol 198-199 ◽  
pp. 1462-1468
Author(s):  
Jing Fang Wang

In this paper, under the conditions of low SNR speech endpoint detection, a feature based on the maximum value of Toeplitz Noise endpoint detection methods. Terms of the method of spectrum from the corresponding sequences with a symmetric Toeplitz matrix constructed using the maximum eigenvalue of the matrix information on the voice signal for dual endpoint detection threshold. New algorithm has been tested to effectively distinguish between speech and noise, low-noise in different environmental conditions has good robustness. With the recent recursive signal analysis methods, the accuracy is higher. The algorithm to calculate the cost of a small, real good, simple and easy to implement.

2011 ◽  
Vol 2-3 ◽  
pp. 135-139
Author(s):  
Jing Jiao Li ◽  
Dong An ◽  
Jiao Wang ◽  
Chao Qun Rong

Speech endpoint detection is one of the key problems in the practical application of speech recognition system. In this paper, speech signal contained chirp is decomposed into several intrinsic mode function (IMF) with the method of ensemble empirical mode decomposition (EEMD). At the same time, it eliminates the modal mix superposition phenomenon which usually comes out in processing speech signal with the algorithm of empirical mode decomposition (EMD). After that, selects IMFs contained major noise through the adaptive algorithm. Finally, the IMFs and speech signal contained chirp are input into the independent component analysis (ICA) and pure voice signal is separated out. The accuracy of speech endpoint detection can be improved in this way. The result shows that the new speech endpoint detection method proposed above is effective, and has strong anti-noises ability, especially suitable for the speech endpoint detection in low SNR.


OALib ◽  
2017 ◽  
Vol 04 (03) ◽  
pp. 1-8
Author(s):  
Jian Wei ◽  
Xiang’e Sun

Author(s):  
A. I. Belousov

The main objective of this paper is to prove a theorem according to which a method of successive elimination of unknowns in the solution of systems of linear equations in the semi-rings with iteration gives the really smallest solution of the system. The proof is based on the graph interpretation of the system and establishes a relationship between the method of sequential elimination of unknowns and the method for calculating a cost matrix of a labeled oriented graph using the method of sequential calculation of cost matrices following the paths of increasing ranks. Along with that, and in terms of preparing for the proof of the main theorem, we consider the following important properties of the closed semi-rings and semi-rings with iteration.We prove the properties of an infinite sum (a supremum of the sequence in natural ordering of an idempotent semi-ring). In particular, the proof of the continuity of the addition operation is much simpler than in the known issues, which is the basis for the well-known algorithm for solving a linear equation in a semi-ring with iteration.Next, we prove a theorem on the closeness of semi-rings with iteration with respect to solutions of the systems of linear equations. We also give a detailed proof of the theorem of the cost matrix of an oriented graph labeled above a semi-ring as an iteration of the matrix of arc labels.The concept of an automaton over a semi-ring is introduced, which, unlike the usual labeled oriented graph, has a distinguished "final" vertex with a zero out-degree.All of the foregoing provides a basis for the proof of the main theorem, in which the concept of an automaton over a semi-ring plays the main role.The article's results are scientifically and methodologically valuable. The proposed proof of the main theorem allows us to relate two alternative methods for calculating the cost matrix of a labeled oriented graph, and the proposed proofs of already known statements can be useful in presenting the elements of the theory of semi-rings that plays an important role in mathematical studies of students majoring in software technologies and theoretical computer science.


2013 ◽  
Vol 845 ◽  
pp. 283-286 ◽  
Author(s):  
Malik Abdul Razzaq Al Saedi ◽  
Mohd Muhridza Yaacob

There is a high risk of insulation system dielectric instability when partial discharge (PD) occurs. Therefore, measurement and monitoring of PD is an important preventive tool to safeguard high-voltage equipment from wanton damage. PD can be detected using optical method to increase the detection threshold and to improve the performance of on-line measurement of PD in noise environment. The PD emitted energy as acoustic emission. We can use this emitted energy to detect PD signal. The best method to detect PD in power transformer is by using acoustic emission. Optical sensor has some advantages such as; high sensitivity, more accuracy small size. Furthermore, in on-site measurements and laboratory experiments, it isoptical methodthat gives very moderate signal attenuations. This paper reviews the available PD detection methods (involving high voltage equipment) such as; acoustic detection and optical detection. The advantages and disadvantages of each method have been explored and compared. The review suggests that optical detection techniques provide many advantages from the consideration of accuracy and suitability for the applications when compared to other techniques.


Author(s):  
S. Abijah Roseline ◽  
S. Geetha

Malware is the most serious security threat, which possibly targets billions of devices like personal computers, smartphones, etc. across the world. Malware classification and detection is a challenging task due to the targeted, zero-day, and stealthy nature of advanced and new malwares. The traditional signature detection methods like antivirus software were effective for detecting known malwares. At present, there are various solutions for detection of such unknown malwares employing feature-based machine learning algorithms. Machine learning techniques detect known malwares effectively but are not optimal and show a low accuracy rate for unknown malwares. This chapter explores a novel deep learning model called deep dilated residual network model for malware image classification. The proposed model showed a higher accuracy of 98.50% and 99.14% on Kaggle Malimg and BIG 2015 datasets, respectively. The new malwares can be handled in real-time with minimal human interaction using the proposed deep residual model.


Sign in / Sign up

Export Citation Format

Share Document