Noah Lee

Noah Lee

Machine learning researcher at MIT Lincoln Laboratory; MSCE student at Texas A&M.

Noah Lee

Hi! I’m Noah Lee, a machine learning researcher at MIT Lincoln Laboratory in Lexington, MA. I work at the intersection of computer vision, foundation models, and multi-sensor imaging.

My research focuses on applying large multimodal foundation models to challenging real-world sensing problems:

  • Vision-language models (VLMs) for few-shot detection — benchmarking foundation models on satellite hyperspectral imagery under limited data, extending them to open-set object detection in X-ray imagery, and analyzing scaling laws with few-shot samples.
  • Foundation models for egocentric action recognition — long-term action recognition on body camera footage, with structured generation for pseudo-labeling unlabeled video.
  • 3D object segmentation in CT volumes — pipelines that detect hazardous materials in luggage at security checkpoints, lifting 2D detection into 3D voxel segmentations.
  • Interactive imaging for passenger checkpoint screening — microservice pipelines, finetuned radar anomaly detection, and self-supervised pose models from millimeter-wave radar.

I’m also an MSCE student at Texas A&M University (Fall 2026 – Spring 2028), where I’m working in the Safety Systems AI Lab under Dr. Namgyun Kim on using vision-language models and computer vision to interpret and vectorize construction blueprint PDFs.

Previously, I collaborated with the DAF-MIT AI Accelerator at MIT on pilot performance prediction from multimodal physiological signals, and I’ve worked as a Systems Engineer co-op at The MITRE Corporation.

I received my B.S. in Computer Science from Northeastern University, graduating Summa Cum Laude with a 3.96 GPA.

News

  • [Aug 2026] Starting MSCE at Texas A&M University (Safety Systems AI Lab).
  • [Jul 2024] Received MIT Lincoln Laboratory 2023–2024 Team Award.

Selected Publications

  • N. Lee, P. W. Moore, L. J. Brattain — Scalable Deep Learning for Pilot Performance Analysis Using Multimodal Physiological Time Series, IEEE HPEC 2023.
  • K. Vogt-Lowell, N. Lee, T. Tsiligkaridis, M. Vaillant — Robust Fine-Tuning of Vision-Language Models for Domain Generalization, IEEE HPEC 2023. (*equal contribution)

See the Publications page for the full list.