Ray Ptucha - Featured Faculty 2018

Thanks primarily to advances in deep learning, we are entering what many argue is the 4th industrial revolution, where the line between silicon and carbon or computer and living organisms blur. This revolution is evolving at an exponential pace, with no historical precedent, disrupting almost every industry. Dr. Ptucha loves deep learning theory, and is committed to using this amazing technology such that our children will have a greater quality of life improvement than any prior generation.

Dr. Ptucha is the founder and director of the Machine Intelligence Laboratory (MIL). His research has broken down barriers between vision and language, developed advanced attention and acoustic models, introduced new context-based video understanding architectures, and developed advanced human- computer interfaces. He has learned methods to take multimodal concepts from audio, image, and written text into a common vector space where similar concepts lie near, dissimilar concepts far. The MIL has introduced a technique called Graph-CNN which can apply filtering and pooling operations on high dimensional graphs, enabling deep learning to be applied to generic graph tensors such as chemical structures and LiDAR point clouds. The MIL is using adversarial techniques to develop ground-breaking data augmentation methods where the computer learns statistical bounds of a data distribution, and then fantasizes new samples which a secondary computer program feel are real. These techniques are currently being used to preserve the culture of Seneca, an endangered Native American heritage by performing speech recognition from Seneca elders.

At the root of Dr. Ptucha’s love for engineering is robotics. He has developed an autonomous wheelchair driven by voice commands, is a co-PI on a team that is training manufacturing robots and fork lifts to efficiently and safely exist with humans on warehouse floors, and has been leading the autonomous golf cart project (shown above) for the last five years.

Dr. Ptucha received his Ph.D. from GCCIS in 2013 where upon his gradation he received the honorable best doctoral dissertation award in 2014. He is an NSF Graduate Research Fellow, NVIDIA Deep Learning Instructor, has been awarded over 30 US patents and over 80 peer reviewed publications. He is the Chair of the Rochester Signal Processing Society and is a staunch supporter of FIRST robotics and STEM curriculums.

Ray Ptucha
Assistant Professor
Kate Gleason College of Engineering