Multi-stream target-speech detection and channel fusion

A method comprising:

• receiving a multichannel audio input signal;
• generating a plurality of enhanced target streams based on the received multichannel audio input signal, each enhanced target stream of the plurality of enhanced target streams being generated using a respective speech enhancement technique of a plurality of speech enhancement techniques;
• and
• enhancing a target audio signal associated with the multichannel audio input signal based at least in part on the plurality of enhanced target streams.

An audio processing device, comprising:

• a processing system;
• and
• a memory storing instructions that, when executed by the processing system, causes the audio processing device to:
• receive a multichannel audio input signal;
• generate a plurality of enhanced target streams based on the received multichannel audio input signal, each enhanced target stream of the plurality of enhanced target streams being generated using a respective speech enhancement technique of a plurality of speech enhancement techniques;
• and
• enhance a target audio signal associated with the multichannel audio input signal based at least in part on the plurality of enhanced target streams.