Eleonora Gualdoni

I am a PhD student in Computational Linguistics at the Universitat Pompeu Fabra, Barcelona, supervised by Gemma Boleda.

In my research, I study how humans and models connect language to visual inputs: I am particularly interested in how people refer to the objects they see, navigating their lexicon according to their communicative needs, and in how this lexicon evolved. I work with deep-learning multimodal models, investigating how they deal with language and vision tasks, and exploring how we can encourage human-like behaviors.

My work is part of the ERC project AMORE: "A MOdel of Reference to Entities" (PI: Gemma Boleda). I am also a member of the UPF research group COLT: "COmputational Linguistics and Linguistic Theory".

Check out my article on the blog of the Cognitive Science society!

News

• Abstract accepted at SCiL: Bridging semantics and pragmatics in information-theoretic emergent communication. Eleonora Gualdoni, Mycal Tucker, Roger Levy, Noga Zaslavsky


• I gave a talk at the LINGUAE lab at ENS, Paris: "Systems of object names"


• Paper accepted in the Journal of Memory and Language! What's in a name? A large-scale computational study on how competition between names affects naming variation. Eleonora Gualdoni, Thomas Brochhagen, Andreas Mädebach, Gemma Boleda


• Paper accepted in Science! From language development to language evolution: A unified view of human lexical creativity. Thomas Brochhagen, Gemma Boleda, Eleonora Gualdoni, Yang Xu.


• I am going to LxMLS 2023, the Lisbon Machine Learning Sumer School!


• July 3rd 2023: we are back from REST-CL 2023, our retreat for PhD students! See our pictures here!


• April 2023: our paper Quantifying informativeness of names in visual space has been accepted to CogSci2023!


• March 16th 2023: I gave a talk at MIT, in the CPL meeting: Object names and their dynamicity.


• March 1st 2023: I joined MIT as a visiting researcher in the Computational Psycholinguistics Lab!


• February 27th 2023: New paper accepted at CVPR 2023: Cross-domain image captioning with discriminative fine-tuning.


• January 31st 2023: check out my article on the blog of the Cognitive Science society!


• December 19th 2022: I’ll give two talks at the Deep Learning Barcelona Symposium.


• October 6th 2022: New paper accepted at EMNLP. Communication breakdown: on the low mutual intelligibility between human and neural captioning. We’ll go to Abu Dhabi!


• September 14th 2022: I am going to give two presentations at the conference (Dis)embodiment at the University of Gothenburg


• August 2022: I am visiting the Cognitive Lexicon Lab at the University of Toronto


• We are going to CogSci! See you in ROOM 104C - MTCC North Building on the 28th of July 2022, from 10:20 to 11:40 ET for our oral presentation of Woman or tennis player? Visual typicality and lexical frequency affect variation in object naming, and in Constitution Hall on the 29th of July 2022, from 10:20 to 12:00 ET for our poster presentation of Effects of task and visual context on referring expressions using natural scenes.


• April 14th 2022: we are back from REST-CL our retreat for PhD students! See our pictures here!

Publications

2024

• Bridging semantics and pragmatics in information-theoretic emergent communication. Eleonora Gualdoni, Mycal Tucker, Roger Levy, Noga Zaslavsky. SCiL.

2023

• What’s in a name? A large-scale computational study on how competition between names affects naming variation. Eleonora Gualdoni, Thomas Brochhagen, Andreas Mädebach, Gemma Boleda. Journal of Memory and Language. [Paper] [Repository] [Demo1] [Demo2]


• From language development to language evolution: A unified view of human lexical creativity. Thomas Brochhagen, Gemma Boleda, Eleonora Gualdoni, Yang Xu. Science. [Paper] [Supplementary material] [Repository]


• Quantifying informativeness of names in visual space. Eleonora Gualdoni, Charles Kemp, Yang Xu, Gemma Boleda. 2023. Proceedings of the 45th Annual Conference of the Cognitive Science Society (CogSci). [Paper]


• Run Like a Girl! Sport-Related Gender Bias in Language and Vision. Sophia Harrison, Eleonora Gualdoni, Gemma Boleda. 2023. Findings of ACL. [Paper]


• Cross-domain image captioning with discriminative fine-tuning. Roberto Dessì, Michele Bevilacqua, Eleonora Gualdoni, Nathanaël Rakotonirina, Francesca Franzon & Marco Baroni. 2023. CVPR. [Paper]

2022

• Communication breakdown: on the low mutual intelligibility between human and neural captioning. Roberto Dessì, Eleonora Gualdoni, Francesca Franzon, Gemma Boleda, Marco Baroni. 2022. The 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP). [Paper]


• Woman or tennis player? Visual typicality and lexical frequency affect variation in object naming. Eleonora Gualdoni, Thomas Brochhagen, Andreas Mädebach, Gemma Boleda. 2022. Proceedings of the 44th Annual Conference of the Cognitive Science Society (CogSci). [Paper] [Repository]


• Effects of task and visual context on referring expressions using natural scenes. Andreas Mädebach, Ekaterina Torubarova, Eleonora Gualdoni, Gemma Boleda. 2022. Proceedings of the 44th Annual Conference of the Cognitive Science Society (CogSci). [Paper] [Repository]


• Horse or pony? Visual typicality and lexical frequency affect variability in object naming. Eleonora Gualdoni, Thomas Brochhagen, Andreas Mädebach, Gemma Boleda. 2022. Proceedings of the Society for Computation in Linguistics (SCiL). Vol. 5 , Article 26 [Paper] [Video 15 mins]

2020

• Grounded and Ungrounded Referring Expressions in Human Dialogues: Language Mirrors Different Grounding Conditions. Eleonora Gualdoni, Raffaella Bernardi, Raquel Fernández, Sandro Pezzelle. 2020. CLiC-it 2020. [Paper]


• Be Different to Be Better! A Benchmark to Leverage the Complementarity of Language and Vision. Sandro Pezzelle, Claudio Greco, Greta Gandolfi, Eleonora Gualdoni, Raffaella Bernardi. 2020. Findings of EMNLP. [Paper] [Website] [Repository]

Presentations

2024

• Talk at LINGUAE lab, ENS, Paris: Systems of object names.

2023

• Poster presentation at CogSci 2023: Quantifying informativeness of names in visual space. [Paper]


• Computational Psycholinguistics Lab meeting, MIT: Object names and their dynamicity. Massachusetts Institute of Technology.

2022

• UR-Ling workshop: How do we call things? Insights from a computational study and a new dataset. Universitat Pompeu Fabra.


• Poster presentation at EMNLP 2022: Communication breakdown: on the low mutual intelligibility between human and neural captioning. [Paper]


• 2 talks at the Deep Learning Barcelona Symposium. Communication breakdown: on the low mutual intelligibility between human and neural captioning & What’s in a name? A large-scale computational study on how competition between names affects naming variation.


• 2 oral presentations at the (Dis)embodiment conference at the University of Gothenburg. Woman or tennis player? Visual typicality and lexical frequency affect variation in object naming. & Effects of task and visual context on referring expressions using natural scenes.


• Oral presentation at CogSci 2022: Woman or tennis player? Visual typicality and lexical frequency affect variation in object naming. [Paper] [Repository]


• Poster presentation at CogSci 2022: Effects of task and visual context on referring expressions using natural scenes. [Paper] [Repository]


• Oral presentation at SCiL 2022: Horse or pony? Visual typicality and lexical frequency affect variability in object naming. [Paper] [Video 15 mins]

2021

• Oral presentation at at Hong Kong Polytechnic University: Visual features predict human agreement on object naming, and graded category membership: insights from a language & vision analysis. Best oral presentation award.

2020

• Oral presentation at TExMod 2020: Adjective Ordering: a Compositional Distributional Semantics Approach. Roberto Zamparelli, Nicola Sartorato, Eleonora Gualdoni, Greta Gandolfi.

CV