OpenSpeakerBeam-SS is an independent reimplementation of SpeakerBeam-SS, a real-time target speaker extraction model combining Conv-TasNet and State Space Modeling (S4D). This project is not affiliated with the original paper's authors.