Projects

GANs for Climate Data Generation

I am the lead student of a group of 6 researching the application of conditional Generative Adversarial Networks to the output of earth system models. I am responsible for the maintenance of the repo, onboarding new students, and prototyping a variety of deep learning architectures.

Our work was featured in the NeurIPS 2020 workshop "Tackling Climate Change with Machine Learning" and the 3rd NOAA workshop on "Leveraging AI in Environmental Sciences".

Links:

• NeurIPS 2020 Climate Change AI Workshop Paper

Key Skills:

• Data Visualization
• Deep Learning
• Distributed Training
• Generative Adversarial Networks
• Geographic Data
• Model Prototyping
• Performance Analysis

Object Detection for Archaeology

I am in a group of 3 applying object detection techniques to aerial LiDAR data over archaeological zones of interests. My main responsibility is the data pipeline which involves dataset design, data exploration, and processing the raw LiDAR features into usable input for downstream tasks.

Dashing: Analysis Framework for Performance Counters

Working in a group of 4, I was the lead designer of an analysis and data pipeline for the study of hardware performance counters. This pipeline was capable of processing data from a variety of sources, performing user-defined analysis, and producing a dashboard containing user-defined visualizations.

This work was featured in the 2019 IEEE/ACM International Workshop on Programming and Performance Visualization Tools.