About

Professor Monica Bugallo's research focuses on statistical signal processing, with emphasis on Bayesian analysis, Sequential Monte Carlo methods, Adaptive filtering, and Stochastic optimization. Her work applies these techniques to multiuser communications, smart antenna systems, target tracking, and vehicle positioning and navigation. She is engaged in advancing the theoretical foundations of signal processing and developing practical applications in communication systems and navigation technologies.

Research topics

• Sociology
• Machine Learning
• Computer Science
• Artificial Intelligence
• Pedagogy
• Mathematics education
• Engineering management
• Psychology
• Algorithm
• Engineering
• Mathematics

Selected publications

• Reducing Diffusion Model Memorization with Higher Order Langevin Dynamics

  ArXiv.org · 2026-05-18

  articleOpen access

  Diffusion/score-based models have emerged as powerful generative models, capable of generating high-quality samples that mimic the training data distribution. However, it has been observed that they are prone to reproducing training samples-known as "memorization"-potentially violating copyright and privacy. In this paper, we study the effect of Higher-Order Langevin Dynamics (HOLD) on this phenomenon. HOLD diffusion processes introduce auxiliary variables; if the data variable is interpreted as "position," then the auxiliary variables can be interpreted as "velocity" and "acceleration," depending on the chosen order of the model. They were originally proposed based on the intuition that they regularize the trajectories of the data variable by implicitly imposing additional dynamical constraints. Our work provides, to our knowledge, the first theoretical characterization of the regularization effect of HOLD. Specifically, we show that in HOLD, the dynamics of the data variable are governed by a low-pass-filtered version of the learned score function, with smoothness increasing with the order of HOLD. We then analyze the optimal empirical score and the possibility of distribution collapse. Together, our results explain the mitigation of memorization as the model order increases. Finally, we present an empirical study on real-world data that supports our theory and highlights this distinct advantage of HOLD over standard diffusion in practice.

  PublisherOA PDF

• Reducing Diffusion Model Memorization with Higher Order Langevin Dynamics

  arXiv (Cornell University) · 2026-05-18

  preprintOpen access

  Diffusion/score-based models have emerged as powerful generative models, capable of generating high-quality samples that mimic the training data distribution. However, it has been observed that they are prone to reproducing training samples-known as "memorization"-potentially violating copyright and privacy. In this paper, we study the effect of Higher-Order Langevin Dynamics (HOLD) on this phenomenon. HOLD diffusion processes introduce auxiliary variables; if the data variable is interpreted as "position," then the auxiliary variables can be interpreted as "velocity" and "acceleration," depending on the chosen order of the model. They were originally proposed based on the intuition that they regularize the trajectories of the data variable by implicitly imposing additional dynamical constraints. Our work provides, to our knowledge, the first theoretical characterization of the regularization effect of HOLD. Specifically, we show that in HOLD, the dynamics of the data variable are governed by a low-pass-filtered version of the learned score function, with smoothness increasing with the order of HOLD. We then analyze the optimal empirical score and the possibility of distribution collapse. Together, our results explain the mitigation of memorization as the model order increases. Finally, we present an empirical study on real-world data that supports our theory and highlights this distinct advantage of HOLD over standard diffusion in practice.

  PublisherDOI

• Indoor Uav Tracking and Altitude Estimation Via Multiple Particle Filters with Partial Parameter Sharing

  2026-04-21

  articleSenior author

  Indoor navigation of unmanned aerial vehicles (UAVs) remains a challenging problem due to the lack of reliable satellite positioning and complex spatial constraints. In many applications, the unknown nature of indoor vertical clearances, such as dropped ceilings, ventilation ducts, and doorways, is an added consideration compared to outdoor settings, where vertical freedom is practically limitless. Traditional UAV navigation methods often assume prior knowledge of the ceiling height, which is an oversimplification of real-world scenarios. This leads to suboptimal path planning and collisions, especially for smaller UAVs, which rely on mounted sensor arrays and have considerable restrictions on payload capacities. In this paper, we focus on simultaneously estimating both the UAV position and the vertical clearance for small-scale UAV flights using multiple particle filters (MPFs), with the height of the indoor environment as an unknown parameter shared across some of the filters. The results show that the proposed method provides more accurate results compared to traditional particle filters.

  PublisherDOI

• Critically-Damped Higher-Order Langevin Dynamics for Generative Modeling

  ArXiv.org · 2025-06-26

  preprintOpen accessSenior author

  Denoising diffusion probabilistic models (DDPMs) represent an entirely new class of generative AI methods that have yet to be fully explored. They use Langevin dynamics, represented as stochastic differential equations, to describe a process that transforms data into noise, the forward process, and a process that transforms noise into generated data, the reverse process. Many of these methods utilize auxiliary variables that formulate the data as a ``position" variable, and the auxiliary variables are referred to as ``velocity", ``acceleration", etc. In this sense, it is possible to ``critically damp" the dynamics. Critical damping has been successfully introduced in Critically-Damped Langevin Dynamics (CLD) and Critically-Damped Third-Order Langevin Dynamics (TOLD++), but has not yet been applied to dynamics of arbitrary order. The proposed methodology generalizes Higher-Order Langevin Dynamics (HOLD), a recent state-of-the-art diffusion method, by introducing the concept of critical damping from systems analysis. Similarly to TOLD++, this work proposes an optimal set of hyperparameters in the $n$-dimensional case, where HOLD leaves these to be user defined. Additionally, this work provides closed-form solutions for the mean and covariance of the forward process that greatly simplify its implementation. Experiments are performed on the CIFAR-10 and CelebA-HQ $256 \times 256$ datasets, and validated against the FID metric.

  PublisherOA PDFDOI

• Transdimensional Model Learning With Online Feature Selection Based on Predictive Least Squares

  IEEE Transactions on Signal Processing · 2025-01-01

  article

  Our interest in obtaining predictive models from vast volumes of data has never been greater. Often, we want these models to be as simple as possible but not simpler. Reducing the complexity or size of a model is accomplished by enforcing sparsity in the model’s parameters or structure. In the context of sparse regression, LASSO and its variants have become standard practice, but they rely on the choice of the penalty parameter for desired degrees of sparsity or predictability. In this work, we develop a novel online inference approach for transdimensional model learning with online feature selection, built on the fundamental principles of predictive least squares. The solution path follows the predictive error at each step, prioritizing more predictive features in the model. We provide a variety of examples to analyze the capabilities of the proposed method and evaluate its performance against both standard and recently proposed feature selection methods.

  PublisherDOI

• Fusion Strategies in Multiple Particle Filtering in the Presence of Shared Unknown Static Parameters

  2025-04-06

  articleSenior author

  The multiple particle filter (MPF) was proposed as a way to tackle online hidden states estimation for high-dimensional scenarios, where traditional PFs fail due to the curse of dimensionality. MPF works by partitioning the state into substates and assigning a separate PF to track each substate, while maintaining mutual communication between the employed filters. Like standard PFs, a common problem in MPFs is dealing with unknown static parameters in the model. However, unlike in standard PFs, estimation of these parameters in MPF requires more caution as the parameters can be shared across partitions. Such a scenario calls for fusion of information between partitions and problem becomes even more challenging when the observations in the model are interacting, i.e., when at least one measured data point is a function of more than one hidden state. In this work, we propose general fusion approaches in MPF for the estimation of static parameters shared across partitions. The proposed strategies use weights to quantify the contribution of each partition to the final estimation and can be applied to the general case with interacting observations. We demonstrate the performance of the proposed method on simulated data and showcase its benefits over several competing approaches.

  PublisherDOI

• BOARD # 315: A Customizable Engineering Outreach Program for Elementary through High School Students (EDU/DRL)

  2025-08-21

  articleSenior author
  PublisherDOI

• Critically-Damped Third-Order Langevin Dynamics

  2025-03-12

  articleSenior author

  While systems analysis has been studied for decades in the context of control theory, it has only been recently used to improve the convergence of Denoising Diffusion Probabilistic Models. This work describes a novel improvement to Third-Order Langevin Dynamics (TOLD), a recent diffusion method that performs better than its predecessors. This improvement, abbreviated TOLD++, is carried out by critically damping the TOLD forward transition matrix similarly to Dockhorn’s Critically-Damped Langevin Dynamics (CLD). Specifically, it exploits eigen-analysis of the forward transition matrix to derive the optimal set of dynamics under the original TOLD scheme. TOLD++ is theoretically guaranteed to converge faster than TOLD, and its faster convergence is verified on the Swiss Roll toy dataset and CIFAR-10 dataset according to the FID metric.

  PublisherDOI

• Particle Markov Chain Monte Carlo Approach to Inference in Transient Surface Kinetics

  Journal of Chemical Theory and Computation · 2025-01-02 · 1 citations

  articleOpen accessSenior author

  In this work, we develop a novel Bayesian approach to study the adsorption and desorption of CO onto a Pd(111) surface, a process of great importance in natural sciences. The motivation for this work comes from the recent availability of time-resolved infrared spectroscopy data and the need for model interpretability and uncertainty quantification in chemical processes. The objective is to learn the relevant parameters that characterize the process: coverage with time, rate constants, activation energies, and pre-exponential factors. Our approach consists of three main schemes: (i) a problem design and probabilistic model for the whole system, (ii) a particle Markov chain Monte Carlo sampler to learn the hidden coverages and rate constant parameters, and (iii) two Bayesian formulations to infer the activation energies and pre-exponential factors. The flexibility of the Bayesian framework allows for uncertainty quantification where possible and integration of mathematical constraints in the model to reflect the system physically. We found that our results for the activation energies and pre-exponential factor are in agreement with those reported in the experimental literature, independently, and we provide discussions on the advantages and disadvantages as well as applicability to other systems.

  PublisherOA PDFDOI

• Fast Sparse Learning from Streaming Data with LASSO

  2025-03-12 · 2 citations

  articleSenior author

  In this paper, we propose Online LASSO - a version of LASSO that is configured for streaming data. In standard LASSO, the penalty parameter is typically chosen by cross-validation, a procedure which requires the entire dataset upfront and repeated fitting. The main contribution of this work is in finding an easy and principled choice for the penalty parameter for every incoming data point, in cases where the input features are uncorrelated. The proposed Online LASSO has several benefits: i) it is memory and time efficient ii) it is easy to implement, iii) it does not require an initial batch of data to start, iv) it does not require any tuning (e.g., step size or tolerance), and finally v) it converges to the performance of the optimal predictor and correct selection of features. We demonstrate these capabilities and compare Online LASSO with standard LASSO as well as other adaptive LASSO variations and provide a discussion on their performances.

  PublisherDOI

Recent grants

• E3: Excellence in Engineering Education - A Workshops Series for School Administrators

  NSF · $100k · 2018–2021

• Studying and Evaluating Education, Guidance, Advancement, and Learning in Technology and Engineering

  NSF · $699k · 2017–2022

• CAREER: Sequential Monte Carlo Methods for High Dimensional Systems

  NSF · $400k · 2010–2016

• CIF: Small: Advancing Adaptive Importance Sampling for Signal Processing

  NSF · $498k · 2016–2020

Frequent coauthors

• Petar M. Djurić

  Stony Brook University

  151 shared

• Joaquı́n Mı́guez

  46 shared

• V́ıctor Elvira

  35 shared

• Luca Martino

  Universidad Rey Juan Carlos

  29 shared

• David Luengo

  21 shared

• Yousef El-Laham

  18 shared

• Angela M. Kelly

  17 shared

• Marco Lops

  16 shared

Labs

• Electrical and Computer Engineering LaboratoryPI