We propose a strategy for discriminative training of the ivector extractor in speaker recognition. The original i-vector extractor training was based on the maximum-likelihood generative modeling, where the EM algorithm was used. In our approach, the i-vector extractor parameters are numerically optimized to minimize the discriminative cross-entropy error function. Two versions of the i-vector ...

In a recent breakthrough [CZ15], Chattopadhyay and Zuckerman gave an explicit two-source extractor for min-entropy k ≥ log n for some large enough constant C. However, their extractor only outputs one bit. In this paper, we improve the output of the two-source extractor to k, while the error remains n−Ω(1). Our improvement is obtained by giving a better extractor for (q, t, γ) non-oblivious bit...

The NucliSens Extractor is an automated nucleic acid isolation system based on guanidinium thiocyanate (GuSCN)-silica extraction technology. The system has been validated for the isolation of human immunodeficiency virus (HIV) and hepatitis C virus (HCV) RNAs from human samples in combination with nucleic acid sequence-based amplification- and reverse transcription-PCR-based methods. We evaluat...

Abstract A construction of Bourgain [Bou05] gave the first 2-source extractor to break the min-entropy rate 1/2 barrier. In this note, we write an exposition of his result, giving a high level way to view his extractor construction. We also include a proof of a generalization of Vazirani’s XOR lemma that seems interesting in its own right, and an argument (due to Boaz Barak) that shows that any...

This paper reviews our on going e orts on designing a real time vision system The system consists of a real time feature extractor and a relaxation network The lter system is capable of performing multiple D non separable lters multi resolution decomposition and steerable transform The relaxation network is capable of performing various relaxation di usion and minimization operations The paper ...

We propose a new customizable tool, Language Independent Feature Extractor (LIFE), which models the inherent patterns of any language and extracts relevant features of the language. There are two contributions of this work: (1) no labeled data is necessary to train LIFE (It works when a sufficient number of unlabeled documents are given), and (2) LIFE is designed to be applicable to any languag...

Fuzzy extractors are a powerful tool to extract randomness from noisy data. A fuzzy extractor can extract randomness only if the source data is discrete while in practice source data is continuous. Using quantizers to transform continuous data into discrete data is a commonly used solution. However, as far as we know no study has been made of the effect of the quantization strategy on the perfo...