UnifyID AI fellowship day

Fall 2019 edition

When: Friday, December 6th, 2019

Where: UnifyID, 603 Jefferson Ave, Redwood City, 94063


While there have been many siloed attempts made by the machine learning community to model Bipedal Gait motion-sensor data, these were done with small datasets oft collected in restricted academic environs. Through this fellowship, we at UnifyID introduced the ImageNet moment for human gait analysis via ‘Project GaitNet’, the first planet-sized motion sensor based human bipedal gait dataset ever curated.


3:00-3:10: Welcome Speech, John Whaley, CEO, UnifyID

3:10-3:20: Fellowship Overview, Vinay Prabhu, Chief Scientist, UnifyID

3:20-4:00: Fellowship Lightning Talks Presented by the fellows

4:00-5:30: Invited Machine Learning Talks Presented by the industry speakers:

5:30-6:30: Happy Hour and Networking

Fellowship theme: GaitNet

Fall cohort fellows:

Fellowship project topics:

  1. SoTA architectures for the GaitNet dataset
  2. Binarized neural networks
  3. Overcoming catastrophic forgetting in CNNs
  4. Quaternion neural networks & Rotation equivariance
  5. Deep-generative models for Gait biometrics


Blake Lemoine

Senior Software Engineer, Google* * This talk is being delivered as a US delegate to the International Standards Organization and not as a Google employee

Talk title: AI Data Ethics

Rob Powers

Motion Scientist, Apple

Talk title: Passively identifying and tracking involuntary movements of Parkinson’s Disease

Saurabh Shintre

Sr. Principal Researcher, NortonLifelock

Talk title: Deepfakes. What’s next?

Subutai Ahmad

VP Research, Numenta

Talk title: Sparsity in the neocortex, and its implications for machine learning