Click on a publication title for more information.
Published in arXiv, 2025
We introduce derive analytic approximations to the moments of the Projected Normal distribution, and introduce a new family of distributions that generalizes the Projected Normal distribution.
Recommended citation: Herrera-Esposito, D.; Burge, J (2025). "Supervised Quadratic Feature Analysis: An information geometry approach to dimensionality reduction." arXiv.
Download Paper | Data and code
Published in arXiv, 2025
In this paper we propose a novel method for supervised dimensionality reduction that maximizes Fisher-Rao information geometry distances between the class-conditional distributions. This maximizes quadratic discriminability between classes, finding features that are useful to use in quadratic discriminant analysis (QDA).
Recommended citation: Herrera-Esposito, D.; Burge, J (2025). "Supervised Quadratic Feature Analysis: An information geometry approach to dimensionality reduction." arXiv.
Download Paper | Data and code
Published in Journal of Neuroscience, 2024
In this paper we trained ideal observer models to estimate 3D motion from naturalistic binocular video clips. We show that the resulting ideal observer exhibit non-obvious behaviors that are similar to those reported in human psychophysics.
Recommended citation: Herrera-Esposito, D.; Burge, J (2024). "Optimal estimation of local motion-in-depth with naturalistic stimuli" Journal of Neuroscience.
Download Paper | Data and code
Published in Eurosurveillance, 2023
In this paper we did a systematic meta-analysis of the SARS-CoV-2 serological literature to collect a dataset on how assay sensitivity changes with time after infection. We report that assay characterisics are an essential determinant of sensitivity dynamics, and discuss the potential biases that this introduces in the literature.
Recommended citation: Nana Owusu-Boaitey, Timothy W Russell, Gideon Meyerowitz-Katz, Andrew T Levin, Daniel Herrera-Esposito (2023). " Dynamics of SARS-CoV-2 seroassay sensitivity: a systematic review and modeling study ." Eurosurveillance. 28(21).
Download Paper | Data and code
Published in Journal of Theoretical Biology, 2022
In this paper we modeled the interaction between non-pharmaceutical interventions for SARS-CoV-2 (e.g. contact tracing) and the number of infected individuals in a population. We find connections between the resulting dynamics and the Allee effect from ecology, and discuss the implications for epidemic management.
Recommended citation: Arim, Matı́as; Herrera-Esposito, D.; Bermolen, Paola; Cabana, Álvaro; Fariello, Marı́a Inés; Lima, Mauricio; Romero, Héctor (2022) . " Contact tracing-induced Allee effect in disease dynamics" Journal of Theoretical Biology. 542.
Download Paper | Data and code
Published in BMJ Global Health, 2022
In this paper, COVID-19 fatality is assessed in developing countries using the golden standard of seroprevalence studies meta-analysis. We find that the infection fatality rate is roughly 2 times higher than in high-income countries. This paper was featured in Nature News, Forbes, Quartz, and other news media.
Recommended citation: Andrew T Levin, Nana Owusu-Boaitey, Sierra Pugh, Bailey K Fosdick, Anthony B Zwi, Anup Malani, Satej Soman, Lonni Besançon, Ilya Kashnitsky, Sachin Ganesh, Aloysius McLaughlin, Gayeong Song, Rine Uhm, Daniel Herrera-Esposito, Gustavo de Los Campos, Ana Carolina Pecanha Peçanha Antonio, Enyew Birru Tadese, Gideon Meyerowitz-Katz (2022). "Assesing the burden of COVID-19 in developing countries: systematic review, meta-analysis and public policy implications." BMJ Global Health. 7.
Download Paper | Data and code
Published in BMC Infectious Diseases, 2022
In this paper we report the age-stratified proportion of SARS-CoV-2 infections that result in severe and critical disease. For this we use the gold standard of serological survey meta-analysis, and collected a unique dataset on hospitalizations and ICU admissions from publich health authorities around the world. We show that the risk of severe and critical disease increases exponentially with age, but much less steeply than the risk of fatal illness. This work has been cited in media outlets and in a United States House of Representatives hearing about school policies during COVID-19.
Recommended citation: Herrera-Esposito, D.; de los Campos, Gustavo. " Age-specific rate of severe and critical SARS-CoV-2 infections estimated with multi-country seroprevalence studies. " BMC Infectious Diseases. 22,311.
Download Paper | Data and code
Published in Vision Research, 2021
In this paper we elaborate on out previous psychophysical findings that higher-order texture statistics are a secondary segmentation cue for humans, compared to spectral statistics. We analyze the contribution of both types of statistics to a natural image segmentation task and find that higher-order statistics are redundant with spectral statistics. We argue that this might explain their secondary role in human perception.
Recommended citation: Herrera-Esposito, D.; Gómez-Sena, L; Coen-Cagli, R. (2021). "Redundancy between spectral and higher-order texture statistics for natural image segmentation." Vision Research. 187
Download Paper | Data and code
Published in Journal of Vision, 2021
In this paper we use psychophysics experiments with Portilla-Simoncelli textures to show that texture segmentation affects contextual modulation in peripheral vision. This flexible contextual modulation cannot be accounted for by popular pooling models of crowding. We probe the relevance of different texture properties to contextual modulation, providing guidance for studying the neural mechanisms of flexible contextual modulation.
Recommended citation: Herrera-Esposito, D.; Coen-Cagli, R.; Gómez-Sena, L; (2021). "Flexible contextual modulation of naturalistic texture perception in peripheral vision" Journal of Vision.
Download Paper | Data and code
Published in Archives of Biochemistry and Biophysics, 2017
Tyrosine nitration is an oxidative post-translational modification that can occur in proteins associated to hydrophobic bio-structures such as membranes and lipoproteins. In this work, we studied tyrosine nitration in membranes using a model system consisting of liposomes with pre-incorporated tyrosine-containing 23 amino acid transmembrane peptides.
Recommended citation: Silvina Bartesaghi, Daniel Herrera, Debora M. Martinez, Ariel Petruk, Veronica Demicheli, Madia Trujillo, Marcelo A. Marti, Dario A. Estrín, Rafael Radi (2017). "Tyrosine oxidation and nitration in transmembrane peptides is connected to lipid peroxidation " Archives of Biochemistry and Biophysics. 622.
Download Paper