WIDEBAND SPEECH COMPRESSION

Research

Supervisor: Prof. Lajos Hanzo

Motivation

In the past decades, there has been an explosion of interest and activity in the area of speech compression. This allows a considerable decrease in the bit rate for transmitting speech on the telephone bandwidth(200-3400 Hz). Several international standards have also been adopted. The most recent standards are the G729 8 kbit/s speech codec with its reduced complexity version Annex A and silence compression scheme version Annex B.

Most work on speech coding today is based on telephone-bandwidth speech, nominally limited to about 3.2 kHz and sampled at a rate of 8 kHz. Wideband speech coding is of increasing interest today and is intended for speech or audio signals of 7 kHz, sampled at 16 kHz. For all of us who grew up with the limited bandwidth built into the conventional telephone, dated way back to the first transcontinental telephone service between New York and San Francisco in 1915, wideband is a new and refreshing experience. This improved quality of wideband speech, as exemplified in the Speech Demo section, will find its way to emerging applications such as teleconferencing, multimedia services and high-definition televisions.

Figure 1 shows a typical energy spectrum of a wideband voiced speech signal. Compared to narrowband telephone speech, the low-frequency enhancement from 50 to 200 Hz contributes to increased naturalness, presence and comfort. The high frequency extension from 3400 to 7000 Hz provides better fricative differentation(for example, s versus f) and therefore higher intelligibility. The naturalness of wideband speech is a significant feature for extended telecommunications processes, such as audio teleconferencing and program broadcasting.