Source separation for automatic speech recognition (ASR)

A system comprising:

• a plurality of microphones;
• a memory;
• a speech recognition engine to recognize speech according to an Automatic Speech Recognition (ASR) process;
• and
• a processor configured to process a plurality of input signals based on an error input, the plurality of input signals representing an acoustic input from the plurality of microphones, and to provide a plurality of output signals configured for the Automatic Speech Recognition (ASR) process at the speech recognition engine,
• wherein the error input is based on feedback from the speech recognition engine, the error input representing an error of the speech recognition engine based on a plurality of previous output signals generated by the processor and provided to the speech recognition engine,
• wherein the processor is configured to generate the plurality of output signals by processing the plurality of input signals to separate the acoustic input into a plurality of source-separated acoustic signals corresponding to a plurality of acoustic sources, respectively, wherein the plurality of output signals represent the plurality of source-separated acoustic signals, respectively.