The re-introduction of Neural Network models for speech recognition, language processing and dialog modeling have dramatically changed how we view and approach these problems. In this talk, I will introduce the application of neural networks for start-of-the-art spoken language interaction, and present recent work in reinforcement learning (RL) for automated optimization of end-to-end task-oriented dialog systems using using both supervised and unsupervised learning.
Prof. Ian Lane is an Associate Research Professor at Carnegie Mellon University. He leads the speech@sv research laboratory at Carnegie Mellon Silicon Valley, which focuses on the intersection of machine learning, speech interaction and hardware technologies. In addition to his work at Carnegie Mellon University Prof. Lane is also CEO and Founder of capio.ai, a Silicon Valley startup focused on bringing human-quality speech recognition to analytics, transcription and human computer interaction products.