Linear-scale perceptual feature extraction for Speech Bandwidth Extensions

スポンサーリンク

概要

• 論文の詳細を見る

• This paper presents a new method to extract linear-scale perceptual feature as a subsitute of MFCCs for highband (3.4kHz∼) in Speech Bandwidth Extensions(BWE). The feature extraction method is based on the mel-scale constrained Nonnegative Matrix Factorization(NMF), which decompose linear-scale log spectrum into a linear combination of mel-scale latent variables. While MFCCs parametrization contains non-invertible procedures, suggested feature is represented in linear-scale and proper to recover the highband time-domain speech. Experiment results report that suggested feature shows better instrumental performance with narrowband MFCCs than real cepstrum without additional computation.

著者

• Chon Sang

  Applied Acoustics Lab. School Of Electrical Eng. And Computer Science Seoul National University

• LEE Mingu

  Applied Acoustics Lab., School of Electrical Eng. and Computer Science, Seoul National University

• Sung Koeng-Mo

  Applied Acoustics Lab., Institute of New Media and Communications, Department of Electrical Engineering, Seoul National University

• Lee Kuekjae

  Applied Acoustics Lab., Institute of New Media and Communications, Department of Electrical Engineering, Seoul National University

• Chon Sang

  Applied Acoustics Lab., Institute of New Media and Communications, Department of Electrical Engineering, Seoul National University

• Lee Mingu

  Applied Acoustics Lab., Institute of New Media and Communications, Department of Electrical Engineering, Seoul National University

関連論文

• Measurement of Early Reflections in a Room with Five Microphone System
• 波長以下の厚さを持つ皮膜のマッチド・フィールド処理を用いた超音波非破壊検査
• A New Robust Acoustic Crosstalk Cancellation Method with Sum and Difference Filter for 3D Audio System
• Adaptive Dynamic Co-interference Cancellation Algorithm for Wireless LAN(Wireless Communication Technology)
• Linear-scale perceptual feature extraction for Speech Bandwidth Extensions
• Efficient Reconstruction of Speakerphone-Mode Cellular Phone Sound for Application to Sound Quality Assessment
• Novel LMS algorithms based on status categorization

スポンサーリンク