Papers by Maxim Korenevsky
2015 IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU), 2015
Lecture Notes in Computer Science, 2015
The methods of integral equations numerical solving are proposed. These methods combine both dete... more The methods of integral equations numerical solving are proposed. These methods combine both deterministic and statistic operations. Like when using deterministic methods, the problem is reduced to solving a set of algebraic equations. But approximation of the integral by finite sum is performed by means of the Monte Carlo method.
WSEAS Transactions on Systems
Paper presents theory of the Sequential Monte Carlo method and its application for devel- opment ... more Paper presents theory of the Sequential Monte Carlo method and its application for devel- opment of adaptive statistical multidimentional integration algorithms. Theorem of Sequential Monte Carlo convergence is given. Several conventional variance reduction methods are used to develop adaptive integration methods. These methods accumulate data about integrand during execution and use it to ac- celerate convergence of integral estimates. Such an approach provides optimal convergence rates for many important functional classes while retains main merits of conventional Monte Carlo integration methods.
Lecture Notes in Computer Science, 2014
This paper deals with topic segmentation of continuous speech. We propose an online segmentation ... more This paper deals with topic segmentation of continuous speech. We propose an online segmentation method that relies on the information about sentence boundaries obtained from an automatic sentence boundary detection system. We show that using information about sentence boundaries to divide continuous speech into fragments for topic classification provides an increase in classification accuracy of about 25-30%, compared to the method where only a threshold on the number of words is used to divide continuous speech into fragments. The highest average classification F-measure for 5 topics obtained in our experiments is 0.79.
Lecture Notes in Computer Science, 2014
Lecture Notes in Computer Science, 2014
Uploads
Papers by Maxim Korenevsky