About

Miaomiao Zhang is an Assistant Professor in the Department of Electrical and Computer Engineering and the Department of Computer Science at the University of Virginia. She completed her Ph.D. in Computer Science at the University of Utah in 2016 and worked as a postdoctoral associate at the Computer Science and Artificial Intelligence Laboratory at the Massachusetts Institute of Technology. Her research focuses on developing machine learning and AI models at the intersection of mathematics and computer engineering, specifically in the field of medical and biological imaging. Her current work spans generative modeling, image registration, segmentation, and statistical shape analysis, with applications in cardiovascular imaging, neuroimaging, and computer-assisted surgery. She has received several awards for her contributions, including the MICCAI young scientist award in 2014, the NIH Trailblazer R21 Award in 2023, and the NSF CAREER Award in 2023.

Research topics

• Computer Science
• Artificial Intelligence
• Machine Learning
• Computer vision
• Mathematical optimization
• Mathematics
• Algorithm

Selected publications

• Learning Group Actions In Disentangled Latent Image Representations

  2026-03-06

  articleOpen access

  Modeling group actions on latent representations enables controllable transformations of high-dimensional image data. Prior works applying group-theoretic priors or modeling transformations typically operate in the high-dimensional data space, where group actions apply uniformly across the entire input, making it difficult to disentangle the subspace that varies under transformations. While latent-space methods offer greater flexibility, they still require manual partitioning of latent variables into equivariant and invariant subspaces, limiting the ability to robustly learn and operate group actions within the representation space. To address this, we introduce a novel end-to-end framework that for the first time learns group actions on latent image manifolds, automatically discovering transformation-relevant structures without manual intervention. Our method uses learnable binary masks with straight-through estimation to dynamically partition latent representations into transformation-sensitive and invariant components. We formulate this within a unified optimization framework that jointly learns latent disentanglement and group transformation mappings. The framework can be seamlessly integrated with any standard encoder-decoder architecture. We validate our approach on five 2D/3D image datasets, demonstrating its ability to automatically learn disentangled latent factors for group actions in diverse data, while downstream classification tasks confirm the effectiveness of the learned representations. Our code is publicly available at GitHub.

  PublisherOA PDFDOI

• Who Is a Better Matchmaker? Human vs. Algorithmic Judge Assignment in a High-Stakes Startup Competition

  Proceedings of the AAAI Conference on Artificial Intelligence · 2026-03-14

  articleOpen access

  There is growing interest in applying artificial intelligence (AI) to automate and support complex decision-making tasks. However, it remains unclear how algorithms compare to human judgment in contexts requiring semantic understanding and domain expertise. We examine this in the context of the judge assignment problem, matching submissions to suitably qualified judges. Specifically, we tackled this problem at the Harvard President’s Innovation Challenge, the university’s premier venture competition awarding over $500,000 to student and alumni startups. This setting represents a real-world environment where high-quality judge assignment is essential. We developed an AI-based judge assignment algorithm, the Hybrid Lexical-Semantic Similarity Ensemble (HLSE), and deployed it at the competition. We then evaluated its performance against human expert assignments using blinded match-quality scores from judges on 309 judge-venture pairs. Using a Mann-Whitney U statistic-based test, we found no statistically significant difference in assignment quality between the two approaches (AUC=0.48, p=0.40); on average, algorithmic matches were rated 3.90 and manual matches 3.94 on a 5-point scale, where 5 indicates an excellent match. Furthermore, manual assignments that previously required a full week could be automated in several hours by the algorithm during deployment. These results demonstrate that HLSE achieves human-expert-level matching quality while offering greater scalability and efficiency, underscoring the potential of AI-driven solutions to support and enhance human decision-making for judge assignment in high-stakes settings.

  PublisherOA PDFDOI

• Electropulsing-induced β-Mg17Al12 phase transformation to synergy enhanced voltage and efficiency of Mg-air batteries

  Journal of Alloys and Compounds · 2026-01-01

  article
  PublisherDOI

• Improved DENSE-guided deep networks to predict myocardial strain from routine cine magnetic resonance images by accounting for rotation

  Journal of Cardiovascular Magnetic Resonance · 2025-01-01

  articleOpen accessSenior author
  PublisherDOI

• Lab-in-the-loop machine learning for brain-targeting delivery system design

  Cell Biomaterials · 2025-07-02 · 9 citations

  articleOpen access
  PublisherOA PDFDOI

• Deep learning to predict myocardial scar burden and uncertainty quantification

  Journal of Cardiovascular Magnetic Resonance · 2025-01-01 · 2 citations

  articleOpen accessSenior author
  PublisherDOI

• Organizational-performance pay and compensation dispersion

  Journal of Organization Design · 2025-02-26 · 1 citations

  articleOpen accessSenior author

  Abstract To drive organizational performance, managers design compensation packages to incentivize the collective contribution of individuals across the organization. Different pay components across individuals may prompt individual concerns based on their views of fairness: those with equity concerns believe that compensation should reflect individual contributions and thus support differences in pay (dispersion), while those with equality concerns believe that similar compensation would be fairer, thus supporting uniform pay (compression). We document the performance implications of compensation dispersion under organizational-performance pay; in increasingly prevalent compensation forms such as stock options or pooled bonuses, individual compensation consists of a pre-designated share of rewards contingent on organizational performance. We argue that prevailing equality concerns under organizational-performance pay generate opposing and asymmetric responses to dispersion from individuals with high vs. low shares of the rewards, leading to lower overall organizational performance under unequal, heterogeneous share dispersion (vs. equal, homogeneous share compression). Evidence from a controlled experiment with online workers and a supplementary field study of professional eSports athletes validates our predictions. Manipulation of two boundary conditions in the experiment further supports the proposed mechanism, as the observed effects are more pronounced when equality concerns are stronger: (1) when the payment scheme involves reward sharing (vs. not) and (2) when the rewards are presented in a percentage framing (vs. a point framing) that facilitates a stronger perception of reward-sharing.

  PublisherOA PDFDOI

• A method for detecting and recognizing Seal instruments based on TrOCR

  Research Square · 2025-06-09

  preprintOpen access
  PublisherOA PDFDOI

• Multi-object tracking review: retrospective and emerging trend

  Artificial Intelligence Review · 2025-05-07 · 23 citations

  articleOpen access

  Multi-object tracking (MOT) is a critical task involving detecting and continuously tracking multiple objects within a video sequence. It is widely used in various fields, such as autonomous driving and intelligent security. In recent years, deep learning architectures have effectively promoted the development of MOT. However, this task poses significant challenges regarding accuracy due to occlusion/truncation, light variation, camera movement. Researchers have proposed many methods to address these issues to reduce trajectory fragmentation, identity switches, and missing targets. To better understand these advancements, it is essential to categorize the approaches based on their methodologies. This article reviewed the recent development of MOT, divided into Tracking by Detection (TBD) and End-to-End (E2E). By introducing and comparing the two types of tracking algorithms, readers can quickly understand the current development status of MOT. Meanwhile, this review summarizes the links to open-source code of excellent algorithms and common benchmark datasets in the appendix. And provide a unified MOT toolkit that includes evaluation and visualization at https://github.com/guanzhiyu817/MOT-tools . In addition, this review discusses the future directions of MOT, specifically cross-modal reasoning.

  PublisherOA PDFDOI

• IrrMap: A Large-Scale Comprehensive Dataset for Irrigation Method Mapping

  2025-08-03

  articleOpen access

  We introduce IrrMap, the first large-scale dataset (1.1 million patches) for irrigation method mapping across regions. IrrMap consists of multi-resolution satellite imagery from LandSat and Sentinel, along with key auxiliary data such as crop type, land use, and vegetation indices. The dataset spans 1,668,899 farms and 11,443,492 acres across multiple western U.S. states from 2013 to 2023, providing a rich and diverse foundation for irrigation analysis and ensuring geospatial alignment and quality control. The dataset is ML-ready, with standardized 224×224 GeoTIFF patches, the multiple input modalities, carefully chosen train-test-split data, and accompanying dataloaders for seamless deep learning model training and benchmarking in irrigation mapping. The dataset is also accompanied by a complete pipeline for dataset generation, enabling researchers to extend IrrMap to new regions for irrigation data collection or adapt it with minimal effort for other similar applications in agricultural and geospatial analysis. We also analyze the irrigation method distribution across crop groups, spatial irrigation patterns (using Shannon diversity indices), and irrigated area variations for both LandSat and Sentinel, providing insights into regional and resolution-based differences. To promote further exploration, we openly release IrrMap, along with the derived datasets, benchmark models, and pipeline code, through a GitHub repository: https://github.com/Nibir088/IrrMap and Data repository: https://huggingface.co/Nibir/IrrMap, providing comprehensive documentation and implementation details.

  PublisherDOI

Frequent coauthors

• William M. Wells

  17 shared

• Polina Golland

  Massachusetts Institute of Technology

  12 shared

• Alexandra J. Golby

  11 shared

• Frank Preiswerk

  Brigham and Women's Hospital

  11 shared

• Sarah Frisken

  11 shared

• Jie Luo

  Harvard University

  10 shared

• Jiarui Xing

  State Key Laboratory of High Performance Civil Engineering Materials

  10 shared

• Alireza Sedghi

  Queen's University

  8 shared

Labs

• Miaomiao Zhang LabPI

Education

• Ph.D., Computer Science

  University of Utah

• Other

  Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology

Awards & honors

• MICCAI Young Scientist Award (2014)
• Runner-up for the Young Scientist Award in 2016
• Best Paper Award, ISBI 2025
• Best Thematic Paper Award, ML4H 2023
• NIH Trailblazer R21 Award 2023