Centre for

Speech Enhancement Tutorial - Introduction

1.1 Noise types

The terms speech enhancement and speech cleaning properly refer respectively to the improvement in the quality or intelligibility of a speech signal and the reversal of degradations that have corrupted it. In practice however, the two terms are used interchangeably.

In this tutorial we will be concerned with cleaning methods that use only a single microphone signal; other, often more effective, techniques are possible if multiple microphone signals are available. The principal degradations that we are concerned with are:

(a) Additive acoustic noise - such as the noise added to the speech signal when recorded in an environment with noticeable background noise, like in an aircraft cockpit.

(b) Acoustic reverberation - results from the additive effect of multiple reflections of an acoustic signal.

(c) Convolutive channel effects - resulting in an uneven or bandlimited response, can result when the communication channel is not modeled effectively for the channel equalizer to remove the channel impulse response.

(d) Non-linear distortion such as arises from clipping - such as when inappropriate gain is applied at the signal input stage.

(e) Additive broadband electronic noise

(f) Electrical interference

(g) Codec distortion - distortion caused by the coding algorithm due to compression

(h) Distortion introduced by recording apparatus - poor response of microphone.

Previous | Next