Synthetic data and generative models.

Some related work:

https://neurips.cc/virtual/2024/poster/94590

https://waymo.com/research/scenediffuser-city-scale-traffic-simulation-via-a-generative-world-model/

https://waymo.com/research/scenediffuser-efficient-and-controllable-driving-simulation-initialization/

• Research under Dr. Chris Metzler and Computational Sensing Lab on computational imaging and machine learning tasks

• Study hypernetworks and optimized training algorithms for multi-parameter image denoising

• Developed novel methods for 3D imaging and tracking, papers at ICCV, CVPR, IEEE TIP

• Worked on confidential, unreleased project

• Implemented data pipeline to analytics platform using AWS services, Apache Spark, and Hadoop

• Cleaned and transformed data in Scala for ingestion by visualizations and models

• Worked with Dr. Hannah Kerner, NASA Harvest, and JPL ML group to develop machine learning anomaly detection algorithms for targets in Mars Rover images

• Ran experiments and re-engineered visualization code to test algorithms designed to run onboard the rover for color and multi-spectral images