Plenary: A review of on-device fully neural end-to-end speech recognition and synthesis algorithms Post date October 12, 2021 ← Extremely low-bit quantization for Transformers → Soundplus