About

David W. Johnson is a professor in the Department of Computer Science at the University of Maryland, with an appointment also in UMIACS. He earned his Ph.D. from the Massachusetts Institute of Technology in 1992. His research areas include Artificial Intelligence and Robotics, as well as Computer Vision and Machine Perception. He has advised students such as Pedro Sandoval Segura, a Ph.D. candidate. His work has been recognized in the news, notably for the development of the LeafSnap iPhone app, which was mentioned in the New York Times in 2013 and 2009 for its ability to identify trees from photos. He has received grants such as the Honda Initiation Grant and has been involved in various research projects and outreach activities at the University of Maryland.

Research topics

• Computer Science
• Psychology
• Data science
• Cognitive psychology
• Mathematics
• Applied psychology
• Economics
• Statistics
• World Wide Web
• Marketing
• Econometrics
• Business
• Biology
• Social psychology
• Cognitive science
• Neuroscience

Selected publications

• Adaptive, Symmetry-Informed Bayesian Metrology for Precise Quantum Technology Measurements

  Physical Review Letters · 2026-02-18 · 1 citations

  preprintOpen access

  High precision measurements are essential to solve major scientific and technological challenges, from gravitational wave detection to healthcare diagnostics. Quantum sensing delivers greater precision, but an in-depth optimization of measurement procedures has been overlooked. Here, we present a systematic strategy for parameter estimation in the low-data limit that integrates experimental control parameters and natural symmetries. The method is guided by a Bayesian quantifier of precision gain, enabling adaptive optimization tailored to the experiment. We provide general expressions for optimal estimators for any parameter. The strategy's power is demonstrated in a quantum technology experiment, in which ultracold caesium atoms are confined in a micromachined hole in an optical fiber. We find a fivefold reduction in the fractional variance of the estimated parameter, compared to the standard measurement procedure. Equivalently, our strategy achieves a target precision with a third of the data points previously required. Such enhanced device performance and accelerated data collection will be essential for applications in quantum computing, communication, metrology, and the wider quantum technology sector.

  PublisherDOI

• Modeling police officers’ deadly force decisions in an immersive shooting simulator

  2025-06-12

  preprintOpen access

  We used an immersive shooting simulator to examine how race, suspect behavior, and policing scenario shape officers’ deadly force decisions. Officers (N = 659) from the Milwaukee Police Department responded to dynamic video scenarios using realistic handgun responses. Mistaken shootings of unarmed Black suspects were more likely than of White suspects, but only when the suspects behaved non-antagonistically. Cognitive modeling showed this race effect arose not from an initial bias to shoot but from differences in evidence accumulation once the object was visible. Scenario and suspect behavior had the largest overall influence, shaping decisions by altering initial proclivity to shoot. Further analysis suggested that suspect behavior within specific scenarios may partially explain observed race effects. These findings provide a process-level account of deadly force decisions, integrating real-world complexity with psychological theory, and offer a framework for improving research and training around police use-of-force.

  PublisherOA PDFDOI

• Exploiting complex 3D-printed surface structures for portable quantum technologies

  arXiv (Cornell University) · 2025-07-02

  preprintOpen access

  Portable quantum technologies require robust, lightweight apparatus with superior performance. For techniques dependent upon high-vacuum environments, such as atom interferometers and atomic clocks, 3D-printing enables new avenues to tailor in-vacuum gas propagation dynamics. We demonstrate intricate, fine-scale surface patterning of 3D-printed vacuum components to increase the rate at which gas particles collide with the surface. By applying a non-evaporable getter coating for use as a surface pump, we show that the patterned surface pumps gas particles 3.8 times faster than an equivalent flat areas. These patterns can be directly integrated into additively manufactured components, enabling application in close proximity to key experimental regions and contributing to overall mass-reduction. We develop numerical simulations that show good agreement with this result and predict up to a ten-fold increase in pumping rate, for realistic surface structures. Our work has direct applications in enabling passively-pumped portable quantum technologies, but also establishes 3D-printing as a powerful technique for the creation of optimized surface patterning to provide enhanced control over high-vacuum gas dynamics for a broad range of applications.

  PublisherOA PDFDOI

• Additive manufacturing of functionalized atomic vapour cells for next-generation quantum technologies

  2025-03-19

  article

  Atomic vapour cells are an indispensable technique for quantum technologies (QT), but potential improvements are limited by the capacities of conventional manufacturing techniques. Exploiting a 3D-printing technique - digital light processing - we demonstrate an additively manufactured glass vapour cell AM capacities, we demonstrate intricate internal architectures, overprint 2D optoelectronical materials to create integrated sensors and surface functionalisation, while also showing the ability to tailor the optical properties of the AM glass by in-situ growth of gold nanoparticles. The produced cells achieve ultra-high vacuum of 2×10−9 mbar and enable Doppler-free spectroscopy; we demonstrate laser frequency stabilisation as a QT application. These results highlight the transformative role that AM can play for QT in enabling compact, optimised and integrated multi-material components and devices.

  PublisherDOI

• Modeling police officers’ deadly force decisions in an immersive shooting simulator

  2024-08-13 · 3 citations

  preprintOpen accessSenior author

  We introduce a novel framework to understand how race, suspect behavior, and policing scenario impact police officers' decision to shoot. We report four principle results with a sample of police officers from the Milwaukee Police Department (\textit{N}=659) that illustrate the utility of this framework: (1) policing scenario and suspect behavior played important roles in officers' decisions; (2) the effect of race on shooting errors depended on whether suspects behaved in an antagonistic or nonantagonistic way; (3) cognitive modeling showed this effect of race was not due to initial biases to shoot Black suspects but instead due to differences in how evidence was gathered between Black and White suspects; and (4) no credible effects of race were observed on response times. Exploration of the data suggests that the race effect may, in part, be due to behaviors performed by particular suspects in specific scenarios. This work provides a novel method and analytic approach for understanding how officers integrate multiple pieces of information during the decision to shoot and how these different sources of information can impact the decision in different ways at different stages. We emphasize that the current report cannot answer the broad question of ``Are police in general biased?,'' but instead is a means to study how officers make deadly force decisions in specific policing scenarios. This sets the stage for researchers and practitioners to obtain the data necessary for designing effective training interventions.

  PublisherOA PDFDOI

• Additive manufacturing of functionalised atomic vapour cells for next-generation quantum technologies

  Quantum Science and Technology · 2024-10-14 · 5 citations

  articleOpen access

  Abstract Atomic vapour cells are an indispensable tool for quantum technologies (QT), but potential improvements are limited by the capacities of conventional manufacturing techniques. Using an additive manufacturing (AM) technique—vat polymerisation by digital light processing—we demonstrate, for the first time, a 3D-printed glass vapour cell. The exploitation of AM capacities allows intricate internal architectures, overprinting of 2D optoelectronical materials to create integrated sensors and surface functionalisation, while also showing the ability to tailor the optical properties of the AM glass by in-situ growth of gold nanoparticles. The produced cells achieve ultra-high vacuum of 2 × 10 −9 mbar and enable Doppler-free spectroscopy; we demonstrate laser frequency stabilisation as a QT application. These results highlight the transformative role that AM can play for QT in enabling compact, optimised and integrated multi-material components and devices.

  PublisherDOI

• Additive Manufacturing of functionalised atomic vapour cells for next-generation quantum technologies

  arXiv (Cornell University) · 2024-06-21

  preprintOpen access

  Atomic vapour cells are an indispensable tool for quantum technologies (QT), but potential improvements are limited by the capacities of conventional manufacturing methods. Using an additive manufacturing (AM) technique - vat polymerisation by digital light processing - we demonstrate, for the first time, a 3D-printed glass vapour cell. The exploitation of AM capacities allows intricate internal architectures, overprinting of 2D optoelectronical materials to create integrated sensors and surface functionalisation, while also showing the ability to tailor the optical properties of the AM glass by in-situ growth of gold nanoparticles. The produced cells achieve ultra-high vacuum of $2 \times 10^{-9}$ mbar and enable Doppler-free spectroscopy; we demonstrate laser frequency stabilisation as a QT application. These results highlight the transformative role that AM can play for QT in enabling compact, optimised and integrated multi-material components and devices.

  PublisherOA PDFDOI

• Modeling police officers’ deadly force decisions in an immersive shooting simulator

  2024-08-13

  preprintOpen access

  We introduce a novel framework to understand how race, suspect behavior, and policing scenario impact police officers' decision to shoot. We report four principle results with a sample of police officers from the Milwaukee Police Department (\textit{N}=659) that illustrate the utility of this framework: (1) policing scenario and suspect behavior played important roles in officers' decisions; (2) the effect of race on shooting errors depended on whether suspects behaved in an antagonistic or nonantagonistic way; (3) cognitive modeling showed this effect of race was not due to initial biases to shoot Black suspects but instead due to differences in how evidence was gathered between Black and White suspects; and (4) no credible effects of race were observed on response times. Exploration of the data suggests that the race effect may, in part, be due to behaviors performed by particular suspects in specific scenarios. This work provides a novel method and analytic approach for understanding how officers integrate multiple pieces of information during the decision to shoot and how these different sources of information can impact the decision in different ways at different stages. We emphasize that the current report cannot answer the broad question of ``Are police in general biased?,'' but instead is a means to study how officers make deadly force decisions in specific policing scenarios. This sets the stage for researchers and practitioners to obtain the data necessary for designing effective training interventions.

  PublisherDOI

• Evidence for Different Roles of Inhibitory and Prospective Intolerance of Uncertainty During Threat Discrimination Learning

  Collabra Psychology · 2023-01-01 · 2 citations

  articleOpen access1st authorCorresponding

  Uncertainty is a core component of threat and associated learning processes. One methodological factor impacting uncertainty in threat learning paradigms is the threat reinforcement rate, which refers to the proportion of times a cue is reinforced with an aversive stimulus. This study tested the effect of partial vs continuous threat reinforcement on threat / safety discrimination learning, as indexed by skin conductance response (SCR). Using a within-participants design, fifty-nine participants completed a task in which three colored shapes were paired with electric shock at reinforcement schedules of 100% (CS+), 50% (CS+) and 0% (CS-). In addition, the study examined the relationship between the Intolerance of Uncertainty scale (IU) and two subscales – inhibitory and prospective IU – with threat discrimination learning. The data show heightened SCR in the continuous vs partial reinforcement condition to all stimuli, but limited evidence of enhanced discrimination learning. Furthermore, no association was observed between total IU score and threat-safety discrimination. However, using a two-factor model of IU, findings showed higher inhibitory IU and higher prospective IU were associated with diminished and heightened threat discrimination, respectively. These results contribute to a fast-growing literature exploring how the uncertainty inherent to predictors of threat, individual differences in sensitivity to uncertainty, and interactions between these two factors, can shape the acquisition of threat memory.

  PublisherOA PDFDOI

• Understanding Mechanisms Behind Discrimination Using Diffusion Decision Modeling

  2021-03-31 · 2 citations

  preprintOpen accessSenior author

  Past research has documented where discrimination occurs or tested interventions that reduce discrimination. Less is known about how discriminatory behavior emerges and the mechanisms through which successful interventions work. Two studies (N > 4500) apply the Diffusion Decision Model (DDM) to the Judgment Bias Task, a measure of discrimination. In control conditions, participants gave preferential treatment (acceptance to a hypothetical honor society) to physically attractive applicants. DDM analyses revealed participants initially favored attractive candidates and attractiveness was accumulated as evidence of being qualified. Two interventions—raising awareness of bias and asking for more deliberative judgments—reduced discrimination through separate mechanisms. Raising awareness reduced biases in drift rates while increasing deliberation raised decision thresholds. This work offers insight into how discrimination emerges and may aid efforts to develop interventions to lessen discrimination.

  PublisherOA PDFDOI

Frequent coauthors

• David Martin

  30 shared

• John E. Donelson

  Universidade Federal do Rio Grande do Norte

  30 shared

• Alan H. Fairlamb

  University of Dundee

  30 shared

• Fred R. Opperdoes

  de Duve Institute

  30 shared

• Mark D. Adams

  Jackson Laboratory

  30 shared

• Joseph Cesario

  Michigan State University

  28 shared

• Ann Cronin

  26 shared

• Barbara Harris

  26 shared

Awards & honors

• Honda Initiation Grant (2007)