Cross-attention Inspired Selective State Space Models for Target Sound Extraction | Xiaol.x | Podwise