Christopher William Pitts

Albuquerque, NM

I work at Sandia National Laboratories, where my department focuses on work in autonomous sensing and perception. I do algorithm design and implementation for machine learning and target recognition, as well as ATR assurance. My particular areas of expertise include few-shot learning, open-set recognition for neural vision systems, and performance assessments of neural recognition models.

When I’m not working on things I can’t talk about, I enjoy losing board games to my wife, building a programmable smart home, and telling my kids why they should love penguins.

education

University of New Mexico

Class of 2026

PhD, Computer Science

Georgia Institute of Technology

Class of 2020

Master, Computer Science

Brigham Young University

Class of 2017

Bachelor, Computer Science

selected publications

Multinomial feature matching for out-of-distribution detection in synthetic aperture radar

Christopher W. Pitts, Devin White, Trilce Estrada, and Gruia-Catalin Roman

In Automatic Target Recognition XXXVI 2026
National Harbor, Maryland, USA
Received Best Paper Award

Abs Website

Out-of-distribution (OOD) detection is an important part of automatic target recognition (ATR) systems. The capability to reject unknown classes improves reliability and trust in an ATR, and permits the use of otherwise closed-set classifiers where open-set recognition is necessary. In this paper we present multinomial feature matching (MFM), a method for detecting OOD data in the latent feature space of neural classifiers, and apply it to an EfficientNet-B7 model trained on the SAMPLE+ dataset. We show that MFM has efficient and low-overhead runtime characteristics, and that it exhibits a high level of performance when applied to OOD targets in SAMPLE+ and other SAR datasets, including both vehicle targets and clutter. MFM achieves a state-of-the-art area under the receiver operating characteristic (ROC) curve (AUROC) score and false positive rate (FPR) at a 95% true positive rate (TPR) (FPR@95) benchmark on the dataset, outperforming both mainstay benchmarks in OOD detection and contemporary work in the field.
Results from the Application of ODIN to SAR Data, with Mitigation of an Unexpected Multimodality Effect

Christopher Pitts, and Michelle Hummell

Automatic Target Recognition Working Group 78 2024
Redstone Arsenal, Alabama, USA
Rapid Iterative Develop And Deploy: Fast Integration On The U-2 Platform

Todd Pitts, Eric Shields, Henry Qiu, Christopher Pitts, Eva Masingale, and Forest Danford

Seventh Annual Workshop on Naval Applications of Machine Learning 2023
San Diego, California, USA

Website
Seascape: a Due-Diligence Framework for Algorithm Acquisition

Christopher Pitts, Forest Danford, Emily Moore, William Marchetto, Henry Qiu, Leon Ross, and Todd Pitts

In Artificial Intelligence and Machine Learning in Defense Applications IV 2022
Berlin, Germany

PDF Slides Website
Seascape: An Algorithm Benchmarking Framework

Christopher Pitts, Forest Danford, Emily Moore, William Marchetto, Henry Qiu, Leon Ross, and Todd Pitts

Sixth Annual Workshop on Naval Applications of Machine Learning 2022
San Diego, California, USA

Website
The Terminator Has A Bug: AI/ML Assurance

Christopher Pitts

2022
Brigham Young University, Provo, Utah, USA

Slides