Method of automatic speech direction reversal and circuit configuration for implementing the method

Abstract

In a method and a circuit for automatic speech direction reversal a reception signal with variable damping is furnished to a loudspeaker and a signal with a variable damping from a microphone is provided as the transmission signal. The reception signal and microphone signal are continuously classified as speech signal or noise. The one signal which is classified as a speech signal, is damped at a first damping value and the other signal is damped at a second damping value, which is greater than the first. These settings are maintained until a signal is classified as noise. When both signals are classified as speech signals, the preceding damping values are retained. When both signals are classified as noise, both attenuations are set at a third damping value located between the first and second damping values. A transition from the first or second damping value to the third damping value is performed more slowly than a transition from the third damping value to the first or second damping value or a transition from the first to the second damping value, and vice versa.