从参考音频和旋律音频中分离人声，仅用人声送入模型 / Extract vocals from both reference and melody audio before feeding into the model
将合成人声与分离出的伴奏混合输出（需先开启人声分离）/ Mix synthesised vocals with separated accompaniment (requires vocal separation)
从参考音频和旋律音频中分离人声，仅用人声送入模型 / Extract vocals from both reference and melody audio before feeding into the model
将合成人声与分离出的伴奏混合输出（需先开启人声分离）/ Mix synthesised vocals with separated accompaniment (requires vocal separation)