Claudia Soares

Claudia Soares

I use optimization, physics, and probability theory to develop robust, interpretable, and trustworthy learning methods to be applied in the real world.

Posts by Collection

group

publications

Distributed, simple and stable network localization

Published in Signal and Information Processing (GlobalSIP), 2014 IEEE Global Conference on, 2014

We propose a simple, stable and distributed algorithm which directly optimizes the nonconvex maximum likelihood criterion for sensor network localization, with no need to tune any free parameter. We reformulate the problem to obtain a gradient Lipschitz cost; by shifting to this cost function we enable a Majorization-Minimization (MM) approach based on quadratic upper bounds that decouple across nodes; the resulting algorithm happens to be.

Recommended citation: C. Soares, J. Xavier, J. Gomes, "Distributed, simple and stable network localization." Signal and Information Processing (GlobalSIP), 2014 IEEE Global Conference on, 2014. https://doi.org/10.1109/GlobalSIP.2014.7032222

Simple and Fast Convex Relaxation Method for Cooperative Localization in Sensor Networks Using Range Measurements

Published in Signal Processing, IEEE Transactions on, 2015

We address the sensor network localization problem given noisy range measurements between pairs of nodes. We approach the nonconvex maximum-likelihood formulation via a known simple convex relaxation. We exploit its favorable optimization properties to the full to obtain an approach that is completely distributed, has a simple implementation at each node, and capitalizes on an optimal gradient method to attain fast convergence. We offer a.

Recommended citation: C. Soares, J. Xavier, J. Gomes, "Simple and Fast Convex Relaxation Method for Cooperative Localization in Sensor Networks Using Range Measurements." Signal Processing, IEEE Transactions on, 2015. https://doi.org/10.1109/TSP.2015.2454853

Collaborative localization of vehicle formations based on ranges and bearings

Published in 2016 IEEE Third Underwater Communications and Networking Conference (UComms), 2016

We examine the problem of jointly determining the positions of multiple underwater vehicles based on a set of pairwise range and bearing measurements taken over time. This extends prior work on the so-called (static) collaborative localization paradigm where a hybrid approach was proposed for seamless instantaneous fusion (i.e., no time dependence) of range and bearing measurements. To incorporate time we add to the original convexified.

Recommended citation: B. Ferreira, J. Gomes, C. Soares, J. P. Costeira, "Collaborative localization of vehicle formations based on ranges and bearings." 2016 IEEE Third Underwater Communications and Networking Conference (UComms), 2016. https://doi.org/10.1109/UComms.2016.7583426

LocDyn: Robust Distributed Localization for Mobile Underwater Networks

Published in IEEE Journal of Oceanic Engineering, 2017

How do we self-localize large teams of underwater nodes using only noisy range measurements? How do we do it in a distributed way, and incorporating dynamics into the problem? How do we reject outliers and produce trustworthy position estimates? And what if some of the vehicles can measure angular information? The stringent acoustic communication constraints and accuracy needs of our geophysical survey application demand fast.

Recommended citation: Claudia Soares, Joao Gomes, Beatriz Quintino Ferreira, Joao Paulo Costeira, "LocDyn: Robust Distributed Localization for Mobile Underwater Networks." IEEE Journal of Oceanic Engineering, 2017. http://dx.doi.org/10.1109/joe.2017.2736951

Fast distributed MAP inference for large-scale graphical models

Published in IEEE EUROCON 2019 -18th International Conference on Smart Technologies, 2019

Bibtex:

Recommended citation: Claudia Soares, Joao Gomes, "Fast distributed MAP inference for large-scale graphical models." IEEE EUROCON 2019 -18th International Conference on Smart Technologies, 2019.

Understanding the Primary-Specialty Referral Mechanism using Network Science

Published in The International Conference on Complex Networks and Their Applications, 2019

Bibtex:

Recommended citation: Joao Casal da Veiga, Qiwei Han, Claudia Soares, "Understanding the Primary-Specialty Referral Mechanism using Network Science." The International Conference on Complex Networks and Their Applications, 2019.

Fast distributed MAP inference for large-scale graphical models

Published in IEEE EUROCON 2019 -18th International Conference on Smart Technologies, 2019

In every domain of life and society, real-world data gains pull for both a more informed decision making and citizenship. Social and human phenomena carry intricate and unknown dependencies unreachable by traditional machine learning approaches, like regression or classification. How to extract value from large amounts of complex and noisy data? Assuming we know the generative model of our data, inference itself is a combinatorial.

Recommended citation: Claudia Soares, Joao Gomes, "Fast distributed MAP inference for large-scale graphical models." IEEE EUROCON 2019 -18th International Conference on Smart Technologies, 2019. http://dx.doi.org/10.1109/eurocon.2019.8861615

Locating a target from uncertain data: convex supersets based on linear-fractional representations

Published in IEEE EUROCON 2019 -18th International Conference on Smart Technologies, 2019

We address the problem of locating a target by using measurements from a set of sensors called anchors. Each anchor has a known location and measures its distance to the target, the measurement being contaminated with additive noise. The noise vector, which contains the noise realizations across the anchors, is naturally unknown, but we assume it to be drawn from a known bounded uncertainty set.Because.

Recommended citation: Claudia Soares, Joao Xavie, "Locating a target from uncertain data: convex supersets based on linear-fractional representations." IEEE EUROCON 2019 -18th International Conference on Smart Technologies, 2019. http://dx.doi.org/10.1109/eurocon.2019.8861977

Finding Common Image Semantics for Urban Perceived Safety Based on Pairwise Comparisons

Published in 2019 27th European Signal Processing Conference (EUSIPCO), 2019

What influences people’s perception of safety in an urban environment? Does everyone perceive safety the same way or do different people look for different contents in an image, safety-wise? We present a user analysis on a crowd-sourced dataset that contains pairwise comparisons regarding the perceived safety of street imagery from different municipalities in the greater Lisbon area, Portugal. We use state-of-the-art semantic segmentation to extract.

Recommended citation: Gabriel Costa, Claudia Soares, Manuel Marques, "Finding Common Image Semantics for Urban Perceived Safety Based on Pairwise Comparisons." 2019 27th European Signal Processing Conference (EUSIPCO), 2019. http://dx.doi.org/10.23919/eusipco.2019.8903115

Conjunction Data Messages behave as a Poisson Process

Published in Workshop on AI for Spacecraft Longevity, 2021

Bibtex:

Recommended citation: Francisco Caldas, Claudia Soares, Cláudia Nunes, Marta Guimarães, Mariana Filipe, Rodrigo Ventura, "Conjunction Data Messages behave as a Poisson Process." Workshop on AI for Spacecraft Longevity, 2021.

Referral prediction in Healthcare using Graph Neural Networks

Published in GReS – Workshop on Graph Neural Networks for Recommendation and Search. Co-located with the ACM RecSys ’21 conference, 2021

Access paper here

Recommended citation: REGINA DUARTE, QIWEI HAN, CLAUDIA SOARES, "Referral prediction in Healthcare using Graph Neural Networks." GReS – Workshop on Graph Neural Networks for Recommendation and Search. Co-located with the ACM RecSys ’21 conference, 2021. https://europe.naverlabs.com/wp-content/uploads/2021/09/DuarteEtAl2021.pdf

STRONG: Synchronous and asynchronous robust network localization, under non-Gaussian noise

Published in Signal Processing, 2021

Real-world network applications must cope with failing nodes, malicious attacks, or nodes facing corrupted data — data classified as outliers. Our work addresses these concerns in the scope of the sensor network localization problem where, despite the abundance of technical literature, prior research seldom considered outlier data. We propose robust, fast, and distributed network localization algorithms, resilient to high-power noise, but also precise under regular.

Recommended citation: Cláudia Soares, João Gomes, "STRONG: Synchronous and asynchronous robust network localization, under non-Gaussian noise." Signal Processing, 2021. https://www.sciencedirect.com/science/article/pii/S0165168421001043

Clustering of the Blendshape Facial Model

Published in 2021 29th European Signal Processing Conference (EUSIPCO), 2021

Digital human animation relies on high-quality 3D models of the human face-rigs. A face rig must be accurate and, at the same time, fast to compute. One of the most common rigging models is the blendshape model. We present a novel approach for learning the inverse rig parameters at increased accuracy and decreased computational cost at the same time. It is based on a two.

Recommended citation: Stevo Rackovic, Claudia Soares, Dusan Jakovetic, Zoranka Desnica, Relja Ljubobratovic, "Clustering of the Blendshape Facial Model." 2021 29th European Signal Processing Conference (EUSIPCO), 2021. http://dx.doi.org/10.23919/eusipco54536.2021.9616061

Decision Support Models for Predicting and Explaining Airport Passenger Connectivity From Data

Published in IEEE Transactions on Intelligent Transportation Systems, 2022

Predicting if passengers in a connecting flight will lose their connection is paramount for airline profitability. We present novel machine learning-based decision support models for the different stages of connection flight management, namely for strategic, pre-tactical, tactical and post-operations. We predict missed flight connections in an airline’s hub airport using historical data on flights and passengers, and analyse the factors that contribute additively to the.

Recommended citation: Marta Guimaraes, Claudia Soares, Rodrigo Ventura, "Decision Support Models for Predicting and Explaining Airport Passenger Connectivity From Data." IEEE Transactions on Intelligent Transportation Systems, 2022. http://dx.doi.org/10.1109/tits.2022.3147155

A Temporal Fusion Transformer for Long-Term Explainable Prediction of Emergency Department Overcrowding

Published in Machine Learning and Principles and Practice of Knowledge Discovery in Databases, 2023

Emergency Departments (EDs) are a fundamental element of the Portuguese National Health Service, serving as an entry point for users with diverse and very serious medical problems. Due to the inherent characteristics of the ED, forecasting the number of patients using the services is particularly challenging. And a mismatch between the affluence and the number of medical professionals can lead to a decrease in the.

Recommended citation: Francisco Caldas, Claudia Soares, "A Temporal Fusion Transformer for Long-Term Explainable Prediction of Emergency Department Overcrowding." Machine Learning and Principles and Practice of Knowledge Discovery in Databases, 2023.

AI for space traffic management

Published in Journal of Space Safety Engineering, 2023

Morgan Stanley forecasts the space industry to top 1 trillion dollars by 2040. Of these 1 trillion dollars, 1.5 billion dollars are expected to be the contribution of the space situational market alone. Satellite operators are already paying the price of space debris. Current approaches for collision avoidance and space traffic management face serious challenges, mainly: (1) Insufficient data and endangered autonomy of action in.

Recommended citation: Chiara Manfletti, Marta Guimaraes, Claudia Soares, "AI for space traffic management." Journal of Space Safety Engineering, 2023. https://www.sciencedirect.com/science/article/pii/S2468896723000897

One-Shot Initial Orbit Determination in Low-Earth Orbit

Published in Accepted at the IEEE Aerospace Conference, 2023

Bibtex:

Recommended citation: Ricardo Ferreira, Marta Guimaraes, Filipa Valdeira, Claudia Soares, "One-Shot Initial Orbit Determination in Low-Earth Orbit." Accepted at the IEEE Aerospace Conference, 2023.

Taxonomy for Resident Space Objects in LEO

Published in Accepted in the International Astronautical Congress 2023, 2023

Bibtex:

Recommended citation: Marta Guimaraes, Claudia Soares, Chiara Manfletti, "Taxonomy for Resident Space Objects in LEO." Accepted in the International Astronautical Congress 2023, 2023.

A majorization-minimization-based method for nonconvex inverse rig problems in facial animation: algorithm derivation

Published in Optimization Letters, 2023

Abstract Automated methods for facial animation are a necessary tool in the modern industry since the standard blendshape head models consist of hundreds of controllers, and a manual approach is painfully slow. Different solutions have been proposed that produce output in real-time or generalize well for different face topologies. However, all these prior works consider a linear approximation of the blendshape function and hence do.

Recommended citation: Stevo Racković, Cláudia Soares, Dušan Jakovetić, Zoranka Desnica, "A majorization-minimization-based method for nonconvex inverse rig problems in facial animation: algorithm derivation." Optimization Letters, 2023. http://dx.doi.org/10.1007/s11590-023-02012-w

Probabilistic Registration for Gaussian Process Three-Dimensional Shape Modelling in the Presence of Extensive Missing Data

Published in SIAM Journal on Mathematics of Data Science, 2023

We propose a shape fitting/registration method based on a Gaussian processes formulation, suitable for shapes with extensive regions of missing data. Gaussian processes are a proven powerful tool, as they provide a unified setting for shape modelling and fitting. While the existing methods in this area prove to work well for the general case of the human head, when looking at more detailed and deformed.

Recommended citation: Filipa M. Valdeira, Ricardo Ferreira, Alessandra Micheletti, Cláudia Soares, "Probabilistic Registration for Gaussian Process Three-Dimensional Shape Modelling in the Presence of Extensive Missing Data." SIAM Journal on Mathematics of Data Science, 2023. http://dx.doi.org/10.1137/22m1495494

Distributed Solution of the Blendshape Rig Inversion Problem

Published in SIGGRAPH Asia 2023 Technical Communications, 2023

The problem of rig inversion is central in facial animation, but with the increasing complexity of modern blendshape models, execution times increase beyond practically feasible solutions. A possible approach towards a faster solution is clustering, which exploits the spacial nature of the face, leading to a distributed method. In this paper, we go a step further, involving cluster coupling to get more confident estimates of.

Recommended citation: Stevo Racković, Cláudia Soares, Dušan Jakovetić, "Distributed Solution of the Blendshape Rig Inversion Problem." SIGGRAPH Asia 2023 Technical Communications, 2023. http://dx.doi.org/10.1145/3610543.3626166

Causal Inference Explanations for Graph Neural Networks

Published in Causal@UAI2024 Poster, 2024

Explainable Artificial Intelligence has emerged, aiming to enhance the trustworthiness of black box models by devising explanation methods that clarify their inner workings. However, prevalent explanation techniques predominantly leverage correlation and association rather than employing causality, a significant aspect of human comprehension. We propose a novel explanation method grounded in causal inference tailored specifically for Graph Neural Networks. Our approach seeks to illuminate the decision-making.

Recommended citation: Sahil Satish Kumar, Claudia Soares, "Causal Inference Explanations for Graph Neural Networks." Causal@UAI2024 Poster, 2024. https://openreview.net/forum?id=xB99i5yHtm

PeersimGym: An Environment for Solving the Task Offloading Problem with Reinforcement Learning

Published in Machine Learning and Knowledge Discovery in Databases. Applied Data Science Track, 2024

Access paper here

Recommended citation: Frederico Metelo, Cláudia Soares, Stevo Racković, Pedro Ákos Costa, "PeersimGym: An Environment for Solving the Task Offloading Problem with Reinforcement Learning." Machine Learning and Knowledge Discovery in Databases. Applied Data Science Track, 2024. http://dx.doi.org/10.1007/978-3-031-70378-2_3

SCENE-Net V2: Interpretable Multiclass 3D Scene Understanding with Geometric Priors

Published in ICML 2024 Workshop GRaM, 2024

In this paper, we present SCENE-Net V2, a new resource-efficient, textbfgray-box model for multiclass 3D scene understanding. SCENE-Net V2 leverages Group Equivariant Non-Expansive Operators (GENEOs) to incorporate fundamental geometric priors as inductive biases, offering a more transparent alternative to the prevalent black-box models in the domain. This model addresses the limitations of its white-box predecessor, SCENE-Net, by expanding its applicability from pole-like structures to a.

Recommended citation: Diogo Mateus Lavado, Claudia Soares, Alessandra Micheletti, "SCENE-Net V2: Interpretable Multiclass 3D Scene Understanding with Geometric Priors." ICML 2024 Workshop GRaM, 2024. https://openreview.net/forum?id=6pKRn6tttu

Precise and Efficient Orbit Prediction in LEO with Machine Learning using Exogenous Variables

Published in 2024 IEEE Congress on Evolutionary Computation (CEC), 2024

The increasing volume of space objects in Earth's orbit presents a significant challenge for Space Situational Awareness (SSA). And in particular, accurate orbit prediction is crucial to anticipate the position and velocity of space objects, for collision avoidance and space debris mitigation. When performing Orbit Prediction (OP), it is necessary to consider the impact of non-conservative forces, such as atmospheric drag and gravitational perturbations, that.

Recommended citation: Francisco Caldas, Cláudia Soares, "Precise and Efficient Orbit Prediction in LEO with Machine Learning using Exogenous Variables." 2024 IEEE Congress on Evolutionary Computation (CEC), 2024. http://dx.doi.org/10.1109/cec60901.2024.10611996

Machine learning in orbit estimation: A survey

Published in Acta Astronautica, 2024

Since the late 1950s, when the first artificial satellite was launched, the number of Resident Space Objects has steadily increased. It is estimated that around one million objects larger than one cm are currently orbiting the Earth, with only thirty thousand larger than ten cm being tracked. To avert a chain reaction of collisions, known as Kessler Syndrome, it is essential to accurately track and.

Recommended citation: Francisco Caldas, Cláudia Soares, "Machine learning in orbit estimation: A survey." Acta Astronautica, 2024. http://dx.doi.org/10.1016/j.actaastro.2024.03.072

Maximum likelihood localization of a network of moving agents from ranges, bearings and velocity measurements

Published in Signal Processing, 2024

Localization is a fundamental enabler technology for many applications, like vehicular networks, IoT, and even medicine. While Global Navigation Satellite Systems solutions offer great performance, they are unavailable in scenarios like indoor or underwater environments, and, for large networks, the instrumentation cost is prohibitive. We develop a localization algorithm from ranges and bearings, suitable for generic mobile networks. Our algorithm is built on a tight.

Recommended citation: Filipa Valdeira, Cláudia Soares, João Gomes, "Maximum likelihood localization of a network of moving agents from ranges, bearings and velocity measurements." Signal Processing, 2024. http://dx.doi.org/10.1016/j.sigpro.2024.109471

Communication-Efficient Vertical Federated Learning via Compressed Error Feedback

Published in IEEE Transactions on Signal Processing, 2025

Communication overhead is a known bottleneck in federated learning (FL). To address this, lossy compression is commonly used on the information communicated between the server and clients during training. In horizontal FL, where each client holds a subset of the samples, such communication-compressed training methods have recently seen significant progress. However, in their vertical FL counterparts, where each client holds a subset of the features.

Recommended citation: Pedro Valdeira, João Xavier, Cláudia Soares, Yuejie Chi, "Communication-Efficient Vertical Federated Learning via Compressed Error Feedback." IEEE Transactions on Signal Processing, 2025. http://dx.doi.org/10.1109/tsp.2025.3540655

Language Modeling and Retrieval-Augmented Generation for Integration in Space Operation Decision Support Tools

Published in 58th IAA Symposium on Safety, Quality and Knowledge Management in Space Activities, 2025

Access paper here

Recommended citation: Ruben Belo, Claudia Soares, Marta Guimaraes, "Language Modeling and Retrieval-Augmented Generation for Integration in Space Operation Decision Support Tools." 58th IAA Symposium on Safety, Quality and Knowledge Management in Space Activities, 2025. http://dx.doi.org/10.52202/083095-0016

OrbitZoo: Real Orbital Systems Challenges for Reinforcement Learning

Published in NeurIPS 2025 poster, 2025

The increasing number of satellites and orbital debris has made space congestion a critical issue, threatening satellite safety and sustainability. Challenges such as collision avoidance, station-keeping, and orbital maneuvering require advanced techniques to handle dynamic uncertainties and multi-agent interactions. Reinforcement learning (RL) has shown promise in this domain, enabling adaptive, autonomous policies for space operations; however, many existing RL frameworks rely on custom-built environments developed.

Recommended citation: Alexandre Oliveira, Katarina Dyreby, Francisco Miguel Caldas, Claudia Soares, "OrbitZoo: Real Orbital Systems Challenges for Reinforcement Learning." NeurIPS 2025 poster, 2025. https://openreview.net/forum?id=oElWLpkOux

Ranking with Confidence for Large Scale Comparison Data

Published in Proceedings of the 2025 SIAM International Conference on Data Mining (SDM), 2025

In this work, we leverage a generative data model considering comparison noise to develop a fast, precise, and informative ranking algorithm from pairwise comparisons that produces a measure of confidence on each comparison. The problem of ranking a large number of items from noisy and sparse pairwise comparison data arises in diverse applications, like ranking players in online games, document retrieval or ranking human perceptions.

Recommended citation: Filipa Valdeira, Cláudia Soares, "Ranking with Confidence for Large Scale Comparison Data." Proceedings of the 2025 SIAM International Conference on Data Mining (SDM), 2025. http://dx.doi.org/10.1137/1.9781611978520.21

Learning Under Noisy Labels, Spurious Points, and Diverse Structures: TS40K, a 3D Point Cloud Dataset of Rural Terrain and Electrical Transmission Systems

Published in 2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2025

Access paper here

Recommended citation: Diogo Lavado, Ricardo Santos, André Coelho, João Santos, Alessandra Micheletti, Claudia Soares, "Learning Under Noisy Labels, Spurious Points, and Diverse Structures: TS40K, a 3D Point Cloud Dataset of Rural Terrain and Electrical Transmission Systems." 2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2025. http://dx.doi.org/10.1109/wacv61041.2025.00712

Generalizing Trilateration: Approximate Maximum Likelihood Estimator for Initial Orbit Determination in Low Earth Orbit

Published in IEEE Transactions on Aerospace and Electronic Systems, 2025

Access paper here

Recommended citation: Ricardo Ferreira, Filipa Valdeira, Marta Guimarães, Cláudia Soares, "Generalizing Trilateration: Approximate Maximum Likelihood Estimator for Initial Orbit Determination in Low Earth Orbit." IEEE Transactions on Aerospace and Electronic Systems, 2025. http://dx.doi.org/10.1109/taes.2025.3566085

SCENE-Net: Geometric induction for interpretable and low-resource 3D pole detection with Group-Equivariant Non-Expansive Operators

Published in Computer Vision and Image Understanding, 2025

This paper introduces SCENE-Net, a novel low-resource, white-box model that serves as a compelling proof-of-concept for 3D point cloud segmentation. At its core, SCENE-Net employs Group Equivariant Non-Expansive Operators (GENEOs), a mechanism that leverages geometric priors for enhanced object identification. Our contribution extends the theoretical landscape of geometric learning, highlighting the utility of geometric observers as intrinsic biases in analyzing 3D environments. Through empirical testing.

Recommended citation: Diogo Lavado, Alessandra Micheletti, Giovanni Bocchi, Patrizio Frosini, Cláudia Soares, "SCENE-Net: Geometric induction for interpretable and low-resource 3D pole detection with Group-Equivariant Non-Expansive Operators." Computer Vision and Image Understanding, 2025. http://dx.doi.org/10.1016/j.cviu.2025.104531

A Systematic Evaluation of Retrieval-Augmented Generation and Language Models for Space Operations

Published in AI4Space, 2026

The rapid expansion of space activities has led to an unprecedented accumulation of technical documentation, operational guidelines, and scientific literature, creating challenges for timely decision-making in space operations. Effective management in space operations requires tools capable of efficiently processing vast and heterogeneous information sources. This paper systematically evaluates the performance of Retrieval-Augmented Generation (RAG) pipelines, combining Large Language Models (LLMs) with information retrieval techniques for.

Recommended citation: Ruben Catarino Belo, Marta Guimaraes, Claudia Soares, "A Systematic Evaluation of Retrieval-Augmented Generation and Language Models for Space Operations." AI4Space, 2026. https://openreview.net/forum?id=hLv173CNNk

research

students

teaching

Machine Learning

MSc course, NOVA School of Science and Technology, Department of Computer Science, 2022

The MSc machine learning course is a rigorous and comprehensive program designed to equip students with advanced knowledge and skills in machine learning. With a focus on both theoretical foundations and practical applications, this course offers students a deep understanding of the algorithms, models, and techniques used in machine learning.

Deep Learning

MSc course, NOVA School of Science and Technology, Department of Computer Science, 2023

Just a few years ago, there was no abundance of deep tech companies, large and small, where deep learning specialists (mostly with PhDs) develop incredible, practical systems that are widely used. These systems cover a variety of research areas, such as computer vision, natural language processing, and reinforcement learning — topics that we’ll cover in this course. The success of deep learning has even inspired research in theoretical learning theory and mathematics. These tools are already having a major influence on industry and society, revolutionizing how films are made, machines are designed, and playing a larger role in fundamental sciences — from astrophysics to medicine.