In order to effectively approach high dimensional pattern recognition problems, one seeks to understand and exploit any inherent low dimensional structure. Recently, a number of manifold learning algorithms have been motivated by a geometric point of view that models high dimensional data as lying near a low dimensional submanifold of the original space. Our paper has two main goals: (i) to inv...

Call centers handle customer queries from various domains such as computer sales and support, mobile phones, car rental, etc. Each such domain generally has a domain model which is essential to handle customer complaints. These models contain common problem categories, typical customer issues and their solutions, greeting styles. Currently these models are manually created over time. Towards th...

Most of the current state-of-the-art speech recognition systems are based on HMMs which usually use mixture of Gaussian functions as state probability distribution model. It is a common practice to use EM algorithm for Gaussian mixture parameter learning. In this case, the learning is done in a ”blind”, data-driven way without taking into account how the speech signal has been produced and whic...

We describe our ongoing work on data-driven models of the tongue shape. Recording techniques such as EMA and X-ray microbeam track the position of 3–4 pellets on the tongue. Our models allow a realistic reconstruction of the full shape of the tongue with submillimetric accuracy from the location of these pellets, and rapid adaptation of an existing model trained with lots of data from one speak...

In this paper we report on advances regarding our approach to porting an automatic speech recognition system to a new target task. In case there is not enough acoustic data available to allow for thorough estimation of HMM parameters it is impossible to train an appropriate model. The basic idea to overcome this problem is to create a task independent seed model that can cope with all tasks equ...

In this paper we report on recent improvements in the University of Colorado system for the DARPA/NRL Speech in Noisy Environments (SPINE) task. In particular, we describe our efforts on improving acoustic and language modeling for the task and investigate methods for unsupervised speaker and environment adaptation from limited data. We show that the MAPLR adaptation method outperforms single a...

Multi-band speech recognition is powerful in band-limited noise, when the recognizer of the noisy band, which is less reliable, can be given less weight in the recombination process. An accurate decision on which bands can be considered as reliable and which bands are less reliable due to corruption by noise is usually hard to take. In this article, we investigate a maximum-likelihood (ML) appr...

The songs of birds, like human voices, are important elements of their identity. In ornithology, distinguishing the songs of different populations is as vital as identifying morphological and genetic differences. This paper takes a machine learning approach to the task. Using data gathered in Indonesia, the song from different subspecies of Black-naped Orioles and Olive-backed Sunbirds is exami...

We present a new approach to solve the problem of phone segmentation when preparing databases for concatenative Text-to-Speech synthesis. First, we describe the problem and review the state of the art. Then we present some already existing techniques to perform this segmentation and present our approach based on a Regression Tree to perform Boundary Specific Correction of the HMM segmentation. ...

Dialogue interaction with remote interlocutors is a difficult application area for speech recognition technology because of the limited duration of acoustic context available for adaptation, the narrow-band and compressed signal encoding used in telecommunications, high variability of spontaneous speech and the processing time constraints. It is even more difficult in the case of interacting wi...