About
Hello! My name is Hernán and since 2021 I am a Ph.D. student in Computer Science at FaMAF of the National University of Córdoba (UNC) advised by Dr. Luciana Benotti with a doctoral scholarship awarded by the national agency for science (CONICET).
In 2019 I obtained a Master’s degree through the final project “Artificial vision for the automatic recognition, in real time, of urban bus lines” in which artificial vision techniques and image analysis were used with the aim of helping to people with visual impairment.
Currently my work is focused in the study of the intersection between modalities of vision and language, to be triyng to extend and adapt existing visual question answering (VQA) models to help blind people in their daily tasks. In particular, I am interested in the way to integrate optical character recognition with the understanding of referring expressions to be able to read and understand texts in images captured by these people.
I have recently begun to study in more detail the large language models. My last work was to create a tool that helps people outside of the computer science field to detect and expose the biases present in these models. And right now I’m fine-tuning a Bert model on the VizWiz dataset, in order to in a later step predict the skills that these VQA systems need to have to improve their responses.
Things I would like to tell you
- March 2023 - Presentation of work in a poster session at the Latin American Meeting In Artificial Intelligence KHIPU 2023. [Poster]
- Dec 2022 - Grant recipient to attend the KHIPU 2023 event, to be held 6-10 of March of 2023 in Montevideo, Uruguay.
- Aug 2022 - I joined FAIR project under the direction of Dra. Laura Alonso Alemany and Dra. Luciana Benotti, for the development of a tool for the detection of biases in large language models. Project led by Vı́a Libre Foundation, with the support of the initiative Women at the Table para una IA Feminista, the Red A+ as well as the EthicsTechLab Laboratory of the University of Notre Dame and the Heinrich Böll Foundation, Cono Sur Office.
- Jul 2022 - Diversity and Inclusion (D&I) grant recipient to virtual attendance to The 2022 Annual Conference of the North American Chapter of the Association for Computational Linguistics (NAACL 2022).
- Dec 2021 - Coding and training of automatic classifier to obtain a new set of categories of visual questions, by disaggregating the data set for blind people VizWiz-VQA. Final project of the postgraduate course Text Mining. FAMAF, UNC, Argentina. Instructor: Dra. Laura Alonso Alemany. [Github]
- Nov 2021 - Diversity and Inclusion (D&I) grant recipient to virtual attendance to The 2021 Conference on Empirical Methods in Natural Language Processing (EMNLP 2021).
- Jul 2021 - Completion of the course “Aprendizaje automático con datos escasos” taught by Jorge A. Sanchez as part of the ECI: Escuela de Ciencias Informáticas, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires. [Certification]
- Apr 2021 - Diversity and Inclusion (D&I) grant recipient to virtual attendance to The 16th Conference of the European Chapter of the Association for Computational Linguistics (EACL 2021).
- Apr 2021 - Doctoral scholarship granted by CONICET. General Topics, Technological and Social Development (KT1).
- Dec 2020 - Coding and training of conditional Deep Convolutional Generative Adversarial Networks (cDCGANs) model to “Automatic Coloring of images”. Final project of the postgraduate course Neural Networks. FAMAF, UNC, Argentina. Instructor: Dr. Francisco Tamarit. [Report PDF]
- Dec 2020 - Completion of the course “Build Basic Generative Adversarial Networks (GANs)” taught by the Coursera platform. [Certification]
- Sep 2020 - Completion of the course “Natural Language Processing with Classification and Vector Spaces” taught by the Coursera platform. [Certification]
Publications
[ 2024 • 2023 • 2022 • 2021 • 2020 • 2019 ]
2024
-
Exploring stereotypes and biases in language technologies in Latin America.
Hernán Maina , Laura Alonso Alemany , Guido Ivetta , Mariela Rajngewerc , Beatriz Busaniche , Luciana Benotti . Communications of the Association for Computing Machinery (ACM). In Press.
2023
-
Bias assessment for experts in discrimination, not in computer science.
Laura Alonso Alemany , Luciana Benotti , Hernán Maina , Lucía González , Lautaro Martínez , Beatriz Busaniche , Alexia Halvorsen , Amanda Mata Rojo , Mariela Rajngewerc . Proceedings of the First Workshop on Cross-Cultural Considerations in NLP (C3NLP). Published.
[PDF] [Code] [Demo]
-
A methodology to characterize bias and harmful stereotypes in natural language processing in Latin America.
Laura Alonso Alemany , Luciana Benotti , Hernán Maina , Lucía González , Mariela Rajngewerc , Lautaro Martínez , Jorge Sánchez , Mauro Schilman , Guido Ivetta , Alexia Halvorsen , Amanda Mata Rojo , Matías Bordone , Beatriz Busaniche . arXiv.org. Preprint.
[PDF] [Code] [Demo]
2022
-
A tool to overcome technical barriers for bias assessment in human language technologies.
Laura Alonso Alemany , Luciana Benotti , Lucía González , Hernán Maina , Beatriz Busaniche , Alexia Halvorsen , Matías Bordone , Jorge Sánchez . arXiv.org. Preprint.
[PDF]
-
What kinds of errors do reference resolution models make and what can we learn from them?.
Jorge Sánchez , Mauricio Mazuecos , Hernán Maina , Luciana Benotti . Findings of the Association for Computational Linguistics: NAACL 2022. Published.
[PDF] [Video]
-
Automatic multi-modal processing of language and vision to assist people with visual impairments.
Hernán Maina , Luciana Benotti . North American Chapter of the Association for Computational Linguistics Conference: LatinX in AI (LXAI) Research Workshop 2022, Virtual. Published.
[PDF] [Slides] [Poster]
-
An interpretable representation of dialog history in referential visual dialog.
Mauricio Mazuecos , Franco Luque , Jorge Sánchez , Hernán Maina , Thomas Vadora , Luciana Benotti . North American Chapter of the Association for Computational Linguistics Conference: LatinX in AI (LXAI) Research Workshop 2022, Virtual. Published.
[PDF]
2021
-
Region under Discussion for visual dialog.
Mauricio Mazuecos , Franco M. Luque , Jorge Sánchez , Hernán Maina , Thomas Vadora and Luciana Benotti . Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing. Published.
[PDF] [Video] [Code]
2020
-
Stop the Bus!: Computer vision for automatic recognition of urban bus lines.
Maina Hernán J , Sánchez Jorge A . XXI Simposio Argentino de Inteligencia Artificial (ASAI 2020)-JAIIO 49 (Modalidad virtual). Published.
[PDF] [Slides] [Video] [Code]
2019
-
Visión artificial para el reconocimiento automático, en tiempo real, de líneas urbanas de autobuses.
Maina Hernán J , Sanchez Jorge A . Thesis work to obtain Master's Degree in Computer Science. Published.
[PDF] [Slides] [Code]
Email Github Google Scholar Acl Anthology