ISCA Archive ICSLP 2002
ISCA Archive ICSLP 2002

Data-driven vector clustering for low-memory footprint ASR

Karim Filali, Xiao Li, Jeff A. Bilmes

It is important to produce automatic speech recognition (ASR) systems that use as few computational and memory resources as possible, especially in low-memory/low-power environments such as for personal digital assistants. One way to achieve this is through parameter quantization. In this work, we compare a variety of novel subvector clustering procedures for ASR system parameter quantization. Specifically, we look at systematic data-driven subvector selection techniques based on entropy minimization, and compare performance on a 150-word isolated word speech recognition task. While the optimal entropy-minimizing quantization methods are intractable, we show that although several of our heuristic techniques are elaborate in their attempt to approximate the optimal clustering, a simple scalar quantization scheme using separate codebooks performs remarkably well.