Cepstral Amplitude Range Normalization for Noise Robust Speech Recognition(Speech and Hearing)
スポンサーリンク
概要
- 論文の詳細を見る
This paper describes a noise robustness technique that normalizes the cepstral amplitude range in order to remove the influence of additive noise. Additive noise causes speech feature mismatches between testing and training environments and it degrades recognition accuracy in noisy environments. We presume an approximate model that expresses the influence by changing the amplitude range and the DC component in the log-spectra. According to this model, we propose a cepstral amplitude range normalization (CARN) that normalizes the cepstral distance between maximum and minimum values. It can estimate noise robust features without prior knowledge or adaptation. We evaluated its performance in an isolated word recognition task by using the Noisex92 database. Compared with the combinations of conventional methods, the CARN could improve recognition accuracy under various SNR conditions.
- 社団法人電子情報通信学会の論文
- 2004-08-01
著者
-
Hayasaka Noboru
Graduate School Of Engineering Hokkaido University
-
Yoshizawa Shingo
Graduate School Of Engineering Hokkaido University
-
Miyanaga Yoshikazu
Graduate School Of Information Science And Technology Hokkaido University
-
WADA Naoya
Graduate School of Environmental Science, Hokkaido University
-
Wada Naoya
Graduate School Of Engineering Hokkaido University
-
Miyanaga Yoshikazu
Graduate School Of Engineering Hokkaido University
関連論文
- A-20-12 Data Frame Format for OFDM System with Variable FFT Point of Data
- Robust Speech Spectra Restoration against Unspecific Noise Conditions for Pitch Detection
- A-20-9 A Study of Phase and Distance Histogram Compensation for OFDM Blind Modulation Detection in Adaptive OFDM Communication
- Performance evaluation of quasi-cyclic LDPC codes for IEEE802.11n based MIMO-OFDM systems (スマートインフォメディアシステム)
- Tunable Wordlength Architecture for a Low Power Wireless OFDM Demodulator(VLSI Design Technology and CAD)
- VLSI Implementation of a Complete Pipeline MMSE Detector for a 4 × 4 MIMO-OFDM Receiver
- A Flexible Architecture for Digital Signal Processing(VLSI System)
- Impact of overgrazing on seed predation by rodents in the Thar desert, northwestern India
- Noise-Robust Speech Analysis Using Running Spectrum Filtering(Speech and Hearing)
- Cepstral Amplitude Range Normalization for Noise Robust Speech Recognition(Speech and Hearing)
- Acoustic Analysis of Vocal Tract Using Auto-Mesh Generation of Finite Element Modeling(Digital Signal Processing)
- High-Speed Finite Element Computation in 3-D Acoustical Analysis of Vocal Tract
- VLSI Implementation of a Scalable Pipeline MMSE MIMO Detector for a 4 x 4 MIMO-OFDM Receiver
- W-04 NEXUS-the Next Generation e-Learning System-and FPGA Hardware Design Platform(International Session)
- Performance and Complexity of MIMO Detectors for Advanced Wireless Communications Systems
- Connectivity Modeling Analysis in Flight-Path Based Aviation Ad Hoc Networks
- New Error Resilience Technique Using Adaptive FMO and Intra Refresh for H.264 Video Transmission
- Design of Area- and Power-Efficient Pipeline FFT Processors for 8x8 MIMO-OFDM Systems
- Development and Outdoor Evaluation of an Experimental Platform in an 80-MHz Bandwidth 2×2 MIMO-OFDM System in 5.2-GHz Band
- A Noise-Robust Continuous Speech Recognition System Using Block-Based Dynamic Range Adjustment
- A Dynamically Reconfigurable FPGA-Based Pattern Matching Hardware for Subclasses of Regular Expressions
- A Dynamically Reconfigurable FPGA-Based Pattern Matching Hardware for Subclasses of Regular Expressions
- A Low Power Tone Recognition for Automatic Tonal Speech Recognizer
- Low-Power Dynamic MIMO Detection for a 4×4 MIMO-OFDM Receiver
- A VLSI Design of a Tomlinson-Harashima Precoder for MU-MIMO Systems Using Arrayed Pipelined Processing
- A Robust Speech Communication into Smart Info-Media System
- Efficiency Improvement in Dynamic Time Warping Algorithms for Isolated Word Recognition