Speaker Embedding-aware Neural Diarization for Flexible Number of Speakers with Textual Information.