Yan Yan

What they work on

1. Develop unsupervised learning models for re-identifying and tracking people and vehicles across a network of city cameras, processing video in real time across edge devices and the cloud.
2. Protect sensitive surveillance video by combining lightweight cryptography with Trusted Execution Environments (TEEs), secure hardware zones, within a scalable and resilient networked system.
3. Create unsupervised tracing algorithms and proofreading software to reconstruct sparsely labeled Brainbow neurons in super-resolution 3D images of whole Drosophila brains.
4. Build a generative model that synthesizes realistic Bitbow neuron images, and use them to train a model that automatically reconstructs densely trans-Tango/Bitbow-labeled fly brain circuits.
5. Build personalized fall-risk and fall-detection models from smartwatch and hip-phone accelerometer data, using small-sample multi-task and multi-modal deep learning in older adults in assisted living.
6. Test whether generative adversarial networks (GANs) can fill in missing movement data by relating accelerometer signals to full-body Azure Kinect skeleton motion, focusing on limb–core dynamics.

More about the research

Why it matters

Cities could use privacy-preserving camera analytics to improve traffic and public safety while protecting pedestrians' and drivers' images. Neuroscientists could reconstruct brain circuits faster and more accurately, and the methods could apply to other biomedical image analysis. Low-cost, personalized fall detection using a smartwatch and phone could help older adults live independently and reduce medical costs from falls.

Known grants

From NIH and NSF research grants. Labs often have other funding we can't see.

Closest labs

The school

University of Illinois Chicago: 42 labs on PI Pages, $532M a year on research, 359 research doctorates a year.

More about University of Illinois Chicago

Reach out

A start for your email. Fill in the brackets.