ADPCM with a multiquantizer for speech coding

Taniguchi, Tomohiko; Unagami, S.; Iseda, K.; Tominaga, S.

doi:10.1109/49.616

Cited by 9 publications

(2 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Multimode coders have been proposed over a wide variety of bandwidths. Taniguchi et al proposed a multimode ADPCM coder at bit rates between 10 and 35 kbps [94]. Johnson and Taniguchi proposed a multimode CELP algorithm at data rates of 4.0-5.3 kbps in which additional stochastic codevectors are added to the LPC excitation vector when channel conditions are sufficiently good to allow high-quality transmission [55].…”

Section: Embedded and Multimode Codingmentioning

confidence: 99%

Speech Coding: Fundamentals and Applications

Hasegawa‐Johnson

Alwan

2003

Wiley Encyclopedia of Telecommunications

View full text Add to dashboard Cite

In this chapter, we present an overview of the most widely used algorithms, standards, and applications of wideband and narrowband speech coding. Algorithms for speech coding are classified into four broad headings: (1) waveform coding techniques (including PCM, companded PCM, and DPCM), which are typically used for landline telephony, internet telephony, and secure military communications; (2) subband coding, including perceptually transparent multi‐rate and embedded coding which is mainly used for internet and digital audio applications; (3) linear predictive analysis by synthesis coding (LPC‐AS) algorithms, including multipulse LPC, CELP, SELP, VSELP, and low‐delay CELP, which are typically used for digital cellular and telephony; and (4) LPC vocoders, including advanced vocoder algorithms (e.g., MELP, MBE, and PWI) are used for applications such as secure telephony and satellite telephony. Applications in areas such as voiceover IP (VoIP) and digital cellular are emerging and require a speech coder to gracefully adapt to rapidly changing channel conditions—a need that is met by embedded and multirate speech coders associated with joint source‐channel coding algorithms. Measures of speech coder perceptual quality include subjective measures of intelligibility (DRT and DALT) and naturalness (MOS and DAM), as well as objective measures such as segmental SNR, Bark spectral distortion, PSQM, and PESQ. Speech coding standards are set by organizations including the ITU (for landline telephony), MPEG (for multimedia applications), ETSI (for European digital cellular), TIA (for U.S. digital cellular), and DDVPC (for United States military applications).

show abstract

Section: Embedded and Multimode Codingmentioning

confidence: 99%

Speech Coding: Fundamentals and Applications

Hasegawa‐Johnson

Alwan

2003

Wiley Encyclopedia of Telecommunications

View full text Add to dashboard Cite

show abstract

“…The optimum bit-rate is selected as follows. ( 1 ) Comparing the input signal level with the input level threshold then determine the evaluation measure.…”

Section: Optimum Control Of the Mq Systemmentioning

confidence: 99%

An implementation of a variable rate codec based on ADPCM with multi-quantizer (ADPCM-MQ)

Tomita

Iseda

Taniguchi

et al.

IEEE Global Telecommunications Conference and Exhibition. Communications for the Information Age

Self Cite

View full text Add to dashboard Cite

Thispaper discusses an implementation of an experimental variable bit-rate codec (VRC) that processes speech signals at 16 kbps to 48 kbps.In previous papers, we proposed a multiple quantizer scheme that improves the transmission characteristics of low bit-rate ADPCM[I]. We investigated this scheme to achieve variable-rate coding of speech.We propose the combined evaluation method of constant SNR and constant noise to optimum bit-rate control of VRC. This algorithm was implemented with six digital signal processor (DSP) chips. The average bitrate of the prototype codec for conversation through the telephone channel was about 23 kbps, including supplementary information, and a quality better than that specified in G.721 (32kbps ADPCM) was observed.

show abstract