Robustness is one of the most important topics for automatic speech recognition (ASR) in practical applications. Monaural speech separation based on computational auditory scene analysis (CASA) offers a solution to this problem. In this paper, a novel system is presented to separate the monaural speech of two talkers. Gaussian mixture models (GMMs) and vector quantizers (VQs) are used to learn ...

The lateral distribution of Cerenkov light at ground level records important information on the development of the air shower which produces it. We have constructed a Broad Lateral Non-imaging Cerenkov Array (BLANCA) to measure this lateral distribution at the CASA-MIA air shower detector in Dugway, Utah. Together, the arrays can sample the lateral distributions of electrons, muons, and Cerenko...

A neural model is described which uses oscillatory correlation to segregate speech from interfering sound sources. The core of the model is a two-layer neural oscillator network. A sound stream is represented by a synchronized population of oscillators, and different streams are represented by desynchronized oscillator populations. The model has been evaluated using a corpus of speech mixed wit...

An auditory “scene”, composed of overlapping acoustic sources, can be viewed as a complex object whose constituent parts are the individual sources. Pitch is known to be an important cue for auditory scene analysis. In this paper, with the goal of building agents that operate in human environments, we describe a real-time system to identify the presence of one or more voices and compute their p...

The human speech recognition system is superior to machine recognition in many ways, but one of the most dramatic is in its resistance to additive "noise" [1]. The human listener is able to use a variety of types of information to help segregate the sounds from different sound sources. The pioneering work on the basic phenomena of human sound segregation was carried out by Al Bregman, and is re...

This paper describes a new method for computational auditory scene analysis which is based on 1) waveform operators to extract instantaneous frequency (IF), frequency change (FM), and amplitude change (AM) from subband signals, and 2) a voting method into a probability distribution to extract coherency (shared fundamental frequency, shared FM, and shared AM) involved in them. We introduce non-p...