State-of-the-Art Speech Recognition Using Multi-Stream Self-Attention with Dilated 1D Convolutions.

Kyu Jeong Han, Ramon Prieto, Tao Ma

Browse the full ASRU paper archive.