Category: AIR Seminar Series

“CreativeAI: Machine Learning meets 3D Content Creation”, Niloy Mitra (University College London)

Niloy Mitra, University College London Monday, November 12, 2018, 10:30am Networking reception, 10:30-11:00am; Seminar 11:00-12:00pm Kilachand Center, 610 Commonwealth Avenue Boston, MA 02215 “CreativeAI: Machine Learning meets 3D Content Creation”   Abstract: A long-standing goal of Computer Graphics is to create high-quality geometric content for a variety of applications including games, movies, product design, and engineering […]

“‘Does This Vehicle Belong to You?’ Processing the Language of Policing for Improving Police-Community Relations” Dan Jurafsky (Stanford)

Dan Jurafsky, Stanford University Monday, October 22, 2018, 10:30am Networking reception, 10:30-11:00am; Seminar 11:00-12:00pm Kilachand Center, 610 Commonwealth Avenue Boston, MA 02215   “Does This Vehicle Belong to You?” Processing the Language of Policing for Improving Police-Community Relations Abstract: Police body-worn cameras have the potential to play an important role in understanding and improving police-community relations. […]

Improving Face Verification Accuracy Using Hyperplane Similarity

This talk by Mike Jones, senior principal research scientist at Mitsubishi Electric Research Labs, will demonstrate how L2 distance is not the best basis of comparison to use in convolutional neural network (CNN) analysis for face verification and propose the hyperplane similarity as a more appropriate similarity function that is derived from the softmax loss function used to train the network.

Knowledge Transfer for Face Recognition

This talk by Zhengming Ding, a graduate student at Northeastern University, outlines a proposal to build a large-scale face recognizer capable of fighting off the data imbalance difficulty that existing machine learning approaches experience in mimicking human visual intelligence. To seek a more effective general classifier, we develop a novel generative model attempting to synthesize meaningful data for one-shot classes by adapting the data variances from other normal classes.