• Publications
  • Influence
The DET curve in assessment of detection task performance
We introduce the DET Curve as a means of representing performance on detection tasks that involve a tradeoff of error types. Expand
Syllable-based large vocabulary continuous speech recognition
We present one of the first robust LVCSR systems that uses a syllable-level acoustic unit for LV CSR on telephone-bandwidth speech that exceeds the performance of a comparable triphone system in terms of word error rate (WER) and complexity. Expand
Geometric linear discriminant analysis for pattern recognition
We propose an innovative technique, geometric linear discriminant analysis (Geometric LDA), to reduce the complexity of pattern recognition systems by using a linear transformation to lower the dimension of the observation space. Expand
A public domain speech-to-text system
We present a public domain Speech-to-Text system where the core module, the decoder, is state-of-the-art in terms of recognition performance as well as consumption of CPU and memory resources. Expand
Syllable-a promising recognition unit for LVCSR
We present an attempt to model syllable level acoustic information as a viable alternative to the conventional phone level acoustic unit for large vocabulary continuous speech recognition. Expand
Geometric linear discriminant analysis
  • M. Ordowski, G. Meyer
  • Mathematics, Computer Science
  • IEEE International Conference on Acoustics…
  • 7 May 2001
We introduce a geometric approach to linear discriminant analysis (GLDA) that can reduce the number of dimensions from n to m for any number of classes and generate the, "optimal" classifier in a lower dimension. Expand
Geometric linear discriminant analysis for pattern recognition systems
More often than not, the complexity of a pattern recognition system is directly related to the difficulty of the task. As the tasks become more difficult, like recognizing objects in digital imagesExpand