Pierre Dognin is a Senior Research Scientist at IBM Research AI at the T.J. Watson Research Center in NY in the Foundations of Trusted AI Department. He has been a member of the Multimodal Team since its creation in 2015 where the research focus is Machine Learning for Multimodal and Trusted AI.

Pierre Dognin completed his PhD in 2003. While a PhD candidate, he gained research experience at the Speech and Language Processing group at BBN Technologies from 2000 to 2003. His research work in speaker normalization was funded by BBN Technologies. He joined IBM Research in 2005 and has been working ever since on technologies for state of the art automatic speech recognition. His research has been on Deep Learning spanning from acoustic modeling for Speech Recognition, to Vision, Language and Graphs more recently, with a focus on multimodal learning (speech, video, text, graphs).

Specialties

• Acoustic Modeling for Speech Recognition.
• Machine Learning, Deep Learning.
• Multimodal systems.

Research interests

Research interests include Multimodal Learning, Generative and Adversarial Learning, Machine Learning, Deep Learning, Computer Vision, Speech Recognition, and any exciting applications of scientific computing.

• Deep Learning, Machine Learning.
• Speech recognition and Multimodal systems.
• Robustness to noise.
• Generative and Adversarial Models
• Large scale AI