Adaptation of compressed HMM parameters for resource-constrained speech recognition

Recently, we successfully developed and reported a new unsupervised online adaptation technique, which jointly compensates for additive and convolutive distortions with vector Taylor series (JAC/VTS), to adjust (uncompressed) HMMs under acoustically distorted environments [1]. In this paper, we extend that technique to adapt compressed HMMs using JAC/VTS where limited computation and/or memory resources are available for speech recognition (e.g., on mobile devices). Subspace coding (SSC) is developed and used to quantize each dimension of the multivariate Gaussians in the compressed HMMs. Three algorithmic design options are proposed and evaluated that combine SSC with JAC/VTS, where three different types of tradeoffs are made between recognition accuracy and the required computation/memory/storage resources. The strengths and weaknesses of these three options are discussed and shown on the Aurora2 task of noise-robust speech recognition. The first option greatly reduces the storage s...

• Type: Conference
• Year: 2008
• Where: ICASSP
• Authors: Jinyu Li, Li Deng, Dong Yu, Jian Wu, Yifan Gong, Alex Acero
• Keywords: Computation And/or Memory, ICASSP 2008, Recognition Accuracy, Signal Processing, Speech Recognition