BROWSE
Detail
Journal Article
Memory Efficient and Fast Speech Recognition System for LowResource Mobile Devices
Cited 8 time in 
Share 
Share
- Authors
- Hoon Chung, Ik Joo Chung
- Issue Date
- 2006-08
- Citation
- IEEE Transactions on Consumer Electronics, v.52, no.3, pp.792-796
- ISSN
- 0098-3063
- Publisher
- IEEE
- Language
- English
- Type
- Journal Article
- Abstract
- In this paper, we consider practical issues such as memory efficiency and fast decoding to make continuous density hidden Markov model (CDHMM)-based large vocabulary speech recognition system work on resource limited mobile devices. Particularly, we focus on memory efficient acoustic modeling and fast state likelihood computation. The proposed techniques are implemented in a speaker-independent Korean speech recognition system running on a Personal Digital Assistant (PDA) with a 32-bit fixed-point processor operating at 400MHz. The system uses 0.5MB memory for representing 28448 Gaussians and it runs at 2.54×RT without serious degradation of accuracy on 10k phonetically optimized words recognition task domain. © 2006 IEEE.
- KSP Keywords
- Acoustic modeling, Continuous Density Hidden Markov Model(CDHMM), Fast Decoding, Fixed-point, Korean speech, Memory Efficiency, Mobile devices, Personal digital assistant(PDA), Practical issues, Speaker-Independent, Speech recognition system