This is the official repository for Interspeech 2024 paper Text-aware Speech Separation for Multi-talker Keyword Spotting. The implementaion of the front-end model is based on ESPnet, which is currently available here in egs2/librimix/enh_kws1. For the KWS backend, We directly apply the default setup of MDTC from WeKws examples/hey_snips/s0.

I apologize that the email address of the primary author is wrong, which should be [email protected] instead of [email protected]. Feel free to mail to me if you have any question!