Auditory Model Based 2.4kbps Speech CODEC and Its Real-Time Implementation Using Single TMS320C5410 DSP Chip

摘要：

In this paper, both sinusoidal representation and secondorder difference cochlear model are used to extract speech feature. Linear interpolation and non-linear smoothing are both used to raise the quality of synthesized speech. A 2.4kbps low bit-rate speech CODEC algorithm is developed after using vector quantisation (VQ) technology to quantise speech features. A single TMS320C5410 DSP chip is adopted to realize a real-time 2.4kbps auditory model based speech CODEC. The experimental results show that the synthesized speech has MOS3.8 quality.

作者: Xiaoqing Yu Wanggen Wan Lei Ma Ning Wang Daniel P.K. Lun

作者单位: Multimedia Innovation Centre, The Hong Kong Polytechnic University, Kowloon, Hong Kong Centre for Multimedia Signal Processing, Department of Electronic and Information Engineering The Ho

会议类型: 国际会议

会议名称: 8th International Conference on Neural Information Processing(ICONIP 2001)(第八届国际神经信息处理大会)

会议地点: 上海

会议语种:英文

页码: 1509-1514

在线出版日期: 2001-11-14（万方平台首次上网日期，不代表论文的发表时间）