Antonio Almudévar

Mr. Antonio Almudévar

orcid

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
29 August 2022 in Sensors
Verified
Journal Article
Open access
PDF file
AAAntonio Almudévar
Antonio Almudévar
PSPascual Sevillano
Pascual Sevillano
LBLuis Vicente Borruel
Luis Vicente Borruel
JPJavier Preciado-Garbayo
Javier Preciado-Garbayo
AOAlfonso Ortega
Alfonso Ortega
Distributed acoustic sensors (DASs) based on direct-detection Φ-OTDR use the light–matter interaction between light pulses and optical fiber to detect mechanical events in the fiber environment. The signals received in Φ-OTDR come from the coherent interference of the portion of the fiber illuminated by the light pulse. Its high sensitivity to minute phase changes in the fiber results in a severe reduction in the signal to noise ratio in the intensity trace that demands processing techniques be able to isolate events. For this purpose, this paper proposes a method based on Unsupervised Anomaly Detection techniques which make use of concepts from the field of deep learning and allow the removal of much of the noise from the Φ-OTDR signals. The fact that this method is unsupervised means that no human-labeled data are needed for training and only event-free data are used for this purpose. Moreover, this method has been implemented and its performance has been tested with real data showing promising results.

5 June 2025 in arXiv
Verified
Preprint
Open access
PDF file
AAAntonio Almudévar
Antonio Almudévar
JHJosé Miguel Hernández-Lobato
José Miguel Hernández-Lobato
SKSameer Khurana
Sameer Khurana
RMRicard Marxer
Ricard Marxer
AOAlfonso Ortega
Alfonso Ortega
Contrastive losses have been extensively used as a tool for multimodal representation learning. However, it has been empirically observed that their use is not effective to learn an aligned representation space. In this paper, we argue that this phenomenon is caused by the presence of modality-specific information in the representation space. Although some of the most widely used contrastive losses maximize the mutual information between representations of both modalities, they are not designed to remove the modality-specific information. We give a theoretical description of this problem through the lens of the Information Bottleneck Principle. We also empirically analyze how different hyperparameters affect the emergence of this phenomenon in a controlled experimental setup. Finally, we propose a regularization term in the loss function that is derived by means of a variational approximation and aims to increase the representational alignment. We analyze in a set of controlled experiments and real-world applications the advantages of including this regularization term.

29 January 2026 in arXiv
Verified
Preprint
Open access
PDF file
AAAntonio Almudévar
Antonio Almudévar
AOAlfonso Ortega
Alfonso Ortega
Machine unlearning seeks to remove the influence of specific training data from a model, a need driven by privacy regulations and robustness concerns. Existing approaches typically modify model parameters, but such updates can be unstable, computationally costly, and limited by local approximations. We introduce Representation Unlearning, a framework that performs unlearning directly in the model's representation space. Instead of modifying model parameters, we learn a transformation over representations that imposes an information bottleneck: maximizing mutual information with retained data while suppressing information about data to be forgotten. We derive variational surrogates that make this objective tractable and show how they can be instantiated in two practical regimes: when both retain and forget data are available, and in a zero-shot setting where only forget data can be accessed. Experiments across several benchmarks demonstrate that Representation Unlearning achieves more reliable forgetting, better utility retention, and greater computational efficiency than parameter-centric baselines.

29 January 2026 in arXiv
Verified
Preprint
Open access
PDF file
AAAntonio Almudévar
Antonio Almudévar
AOAlfonso Ortega
Alfonso Ortega
We present a unified framework for quantifying the similarity between representations through the lens of \textit{usable information}, offering a rigorous theoretical and empirical synthesis across three key dimensions. First, addressing functional similarity, we establish a formal link between stitching performance and conditional mutual information. We further reveal that stitching is inherently asymmetric, demonstrating that robust functional comparison necessitates a bidirectional analysis rather than a unidirectional mapping. Second, concerning representational similarity, we prove that reconstruction-based metrics and standard tools (e.g., CKA, RSA) act as estimators of usable information under specific constraints. Crucially, we show that similarity is relative to the capacity of the predictive family: representations that appear distinct to a rigid observer may be identical to a more expressive one. Third, we demonstrate that representational similarity is sufficient but not necessary for functional similarity. We unify these concepts through a task-granularity hierarchy: similarity on a complex task guarantees similarity on any coarser derivative, establishing representational similarity as the limit of maximum granularity: input reconstruction.