Beatrice Biancardi
Beatrice Biancardi has a total of 40 co-authors. Here are the top co-authors by number of shared publications:
- Stéphanie Buisine
CESI, Nanterre, France
7 shared publications - Mathieu Chollet
IMT Atlantique, LS2N, UMR CNRS 6004
5 shared publications - Mukesh Barange
CESI LINEACT, Nanterre, France
5 shared publications - Tristan Lannuzel
CESI LINEACT, Nanterre, France
4 shared publications - Chloé Clavel
ALMAnaCH, INRIA Paris, Paris, France
3 shared publications - Alisa Barkar
LS2N IMT Atlantique France
2 shared publications
Latest Publications
- Open Access
- 19 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-14
Virtual Reality for Collective Intelligence: A Multimodal Computational Framework
- 19 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-14
Open AccessVirtual Reality for Collective Intelligence: A Multimodal Computational Framework
- Open Access
- 5 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 156-159
CreaRobot - Improving creativity using a social robot
- 5 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 156-159
Open AccessCreaRobot - Improving creativity using a social robot
- Open Access
- 5 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 136-140
When Coordination Signals Can Mislead: Conversational and Facial Markers of Transactive Functioning in Newly-Formed Virtual-Reality Teams
- 5 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 136-140
Open AccessWhen Coordination Signals Can Mislead: Conversational and Facial Markers of Transactive Functioning in Newly-Formed Virtual-Reality Teams
- 2 September 2026
- Published by Springer Nature in Journal on Multimodal User Interfaces
- Vol. 20 (2) , 217-244
A computational framework for modeling collective intelligence through multimodal interaction in virtual reality
- 2 September 2026
- Published by Springer Nature in Journal on Multimodal User Interfaces
- Vol. 20 (2) , 217-244
A computational framework for modeling collective intelligence through multimodal interaction in virtual reality
- 4 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
First seconds matter : Mangaing first impressions for a more engaging virtual agent
Les premières secondes comptent : gérer les premières impressions pour un agent virtuel plus engageant: Les premiers instants d'une interaction avec un personnage virtuel sont critiques, car les impressions que les utilisateurs se forgent à leur sujet peuvent affecter le reste de l'interaction, en termes d'engagement et de volonté de poursuivre l’interaction. Dans cette thèse, nous présentons un modèle computationnel pour un agent virtuel pour gérer les impressions de chaleur et de compétence (C&C), les deux dimensions fondamentales de la cognition sociale, que se forge l’utilisateur. Le but de l'agent est d'adapter son comportement non verbal en temps réel pendant une interaction, en fonction des réactions non verbales de l'utilisateur. Notre approche part de l'analyse d'un corpus d'interactions humaines afin d'identifier un ensemble de comportements non verbaux suscitant la perception de différents degrés de C&C. Une étude perceptive a examiné comment ces comportements sont perçus lorsqu'ils sont réalisés par un agent virtuel. Un algorithme d'apprentissage par renforcement a été développé pour permettre à l'agent d'apprendre en temps réel les comportements qui donnent la meilleure impression à l'utilisateur, en fonction de son objectif. Les réactions non verbales de l'utilisateur (calculées à partir de signaux de bas niveau telles que les unités d'action du visage) sont utilisées comme récompense pour l'algorithme. Nous avons personnalisé le modèle afin d'adapter les comportements de l'agent dans le but de maximiser l'engagement de l'utilisateur ou ses impressions sur la C&C de l'agent. 2 cas d'utilisation ont été menés afin d'évaluer l'impact d'un agent adaptatif sur les impressions de l’utilisateur et sa perception de l'interaction, par rapport à un agent non adaptatif.
- 4 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
First seconds matter : Mangaing first impressions for a more engaging virtual agent
Les premières secondes comptent : gérer les premières impressions pour un agent virtuel plus engageant: Les premiers instants d'une interaction avec un personnage virtuel sont critiques, car les impressions que les utilisateurs se forgent à leur sujet peuvent affecter le reste de l'interaction, en termes d'engagement et de volonté de poursuivre l’interaction. Dans cette thèse, nous présentons un modèle computationnel pour un agent virtuel pour gérer les impressions de chaleur et de compétence (C&C), les deux dimensions fondamentales de la cognition sociale, que se forge l’utilisateur. Le but de l'agent est d'adapter son comportement non verbal en temps réel pendant une interaction, en fonction des réactions non verbales de l'utilisateur. Notre approche part de l'analyse d'un corpus d'interactions humaines afin d'identifier un ensemble de comportements non verbaux suscitant la perception de différents degrés de C&C. Une étude perceptive a examiné comment ces comportements sont perçus lorsqu'ils sont réalisés par un agent virtuel. Un algorithme d'apprentissage par renforcement a été développé pour permettre à l'agent d'apprendre en temps réel les comportements qui donnent la meilleure impression à l'utilisateur, en fonction de son objectif. Les réactions non verbales de l'utilisateur (calculées à partir de signaux de bas niveau telles que les unités d'action du visage) sont utilisées comme récompense pour l'algorithme. Nous avons personnalisé le modèle afin d'adapter les comportements de l'agent dans le but de maximiser l'engagement de l'utilisateur ou ses impressions sur la C&C de l'agent. 2 cas d'utilisation ont été menés afin d'évaluer l'impact d'un agent adaptatif sur les impressions de l’utilisateur et sa perception de l'interaction, par rapport à un agent non adaptatif.
- Open Access
- 8 September 2025
- Published by Springer Nature in Research Square
How LLMs Assess Public Speaking? Methodology of Explaining LLM Judgments through Linguistic Patterns and Rhetorical Criteria
This study examines how Large Language Models, specifically GPT-4o-mini, evaluate public speaking performances based on textual transcripts. We collect a new annotation of speeches from the 3MT_French dataset and compare GPT-4o-mini annotations to those of an expert across both concrete rhetorical criteria and abstract subjective dimensions. In contrast to the expert, GPT-4o-mini exhibits limited cross-criterion integration and struggles with subjective judgments such as persuasiveness and creativity. Further, we utilise linguistic features in order to interpret provided annotations and demonstrated that GPT-4o-mini relies heavily on surface-level linguistic features and tends to prioritise structural and stylistic markers, while expert annotations reflect broader discourse-level understanding and persuasive intent. These findings highlight the limitations of using LLMs in high-level rhetorical evaluation and suggest the need for hybrid systems that combine model capabilities with theory-driven evaluation criteria. The annotated dataset and code are released to support future work in this direction: https://github.com/abarkar/ How-LLMs-Assess-Public-Speaking-/tree/main
- 8 September 2025
- Published by Springer Nature in Research Square
Open AccessHow LLMs Assess Public Speaking? Methodology of Explaining LLM Judgments through Linguistic Patterns and Rhetorical Criteria
This study examines how Large Language Models, specifically GPT-4o-mini, evaluate public speaking performances based on textual transcripts. We collect a new annotation of speeches from the 3MT_French dataset and compare GPT-4o-mini annotations to those of an expert across both concrete rhetorical criteria and abstract subjective dimensions. In contrast to the expert, GPT-4o-mini exhibits limited cross-criterion integration and struggles with subjective judgments such as persuasiveness and creativity. Further, we utilise linguistic features in order to interpret provided annotations and demonstrated that GPT-4o-mini relies heavily on surface-level linguistic features and tends to prioritise structural and stylistic markers, while expert annotations reflect broader discourse-level understanding and persuasive intent. These findings highlight the limitations of using LLMs in high-level rhetorical evaluation and suggest the need for hybrid systems that combine model capabilities with theory-driven evaluation criteria. The annotated dataset and code are released to support future work in this direction: https://github.com/abarkar/ How-LLMs-Assess-Public-Speaking-/tree/main
- Open Access
- 3 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 129-138
Towards a Dynamic Model of Collective Intelligence: Theoretical Integration, Nonverbal Interaction and Temporality
- 3 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 129-138
Open AccessTowards a Dynamic Model of Collective Intelligence: Theoretical Integration, Nonverbal Interaction and Temporality
- Open Access
The Anatomy of Speech Persuasion: Linguistic Shifts in LLM-Modified Speeches
This study examines how large language models understand the concept of persuasiveness in public speaking by modifying speech transcripts from PhD candidates in the "Ma These en 180 Secondes" competition, using the 3MT French dataset. Our contributions include a novel methodology and an interpretable textual feature set integrating rhetorical devices and discourse markers. We prompt GPT-4o to enhance or diminish persuasiveness and analyze linguistic shifts between original and generated speech in terms of the new features. Results indicate that GPT-4o applies systematic stylistic modifications rather than optimizing persuasiveness in a human-like manner. Notably, it manipulates emotional lexicon and syntactic structures (such as interrogative and exclamatory clauses) to amplify rhetorical impact.
Open AccessThe Anatomy of Speech Persuasion: Linguistic Shifts in LLM-Modified Speeches
This study examines how large language models understand the concept of persuasiveness in public speaking by modifying speech transcripts from PhD candidates in the "Ma These en 180 Secondes" competition, using the 3MT French dataset. Our contributions include a novel methodology and an interpretable textual feature set integrating rhetorical devices and discourse markers. We prompt GPT-4o to enhance or diminish persuasiveness and analyze linguistic shifts between original and generated speech in terms of the new features. Results indicate that GPT-4o applies systematic stylistic modifications rather than optimizing persuasiveness in a human-like manner. Notably, it manipulates emotional lexicon and syntactic structures (such as interrogative and exclamatory clauses) to amplify rhetorical impact.
- 11 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Visualization and Computer Graphics
- Vol. 31 (5) , 3306-3313
The Hidden Face of the Proteus Effect: Deindividuation, Embodiment and Identification
The Proteus effect describes how users of virtual environments adjust their attitudes to match stereotypes associated with their avatar's appearance. While numerous studies have demonstrated this phenomenon's reliability, its underlying processes remain poorly understood. This work investigates deindividuation's hypothesized but unproven role within the Proteus effect. Deindividuated individuals tend to follow situational norms rather than personal ones. Therefore, together with high embodiment and identification processes, deindividuation may lead to a stronger Proteus effect. We present two experimental studies. First, we demonstrated the emergence of the Proteus effect in a real-world academic context: engineering students got better scores in a statistical task when embodying Albert Einstein's avatar compared to a control one. In the second study, we tested the role of deindividuation by manipulating participants' exposure to different identity cues during the task. While we could not find a significant effect of deindividuation on the participants' performance, our results highlight an unexpected pattern, with embodiment as a negative predictor and identification as a positive predictor of performance. These results open avenues for further research on the processes involved in the Proteus effect, particularly those focused on the relation between the avatar and the nature of the task to be performed. All supplemental materials are available at https://osf.io/au3wk/.
- 11 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Visualization and Computer Graphics
- Vol. 31 (5) , 3306-3313
The Hidden Face of the Proteus Effect: Deindividuation, Embodiment and Identification
The Proteus effect describes how users of virtual environments adjust their attitudes to match stereotypes associated with their avatar's appearance. While numerous studies have demonstrated this phenomenon's reliability, its underlying processes remain poorly understood. This work investigates deindividuation's hypothesized but unproven role within the Proteus effect. Deindividuated individuals tend to follow situational norms rather than personal ones. Therefore, together with high embodiment and identification processes, deindividuation may lead to a stronger Proteus effect. We present two experimental studies. First, we demonstrated the emergence of the Proteus effect in a real-world academic context: engineering students got better scores in a statistical task when embodying Albert Einstein's avatar compared to a control one. In the second study, we tested the role of deindividuation by manipulating participants' exposure to different identity cues during the task. While we could not find a significant effect of deindividuation on the participants' performance, our results highlight an unexpected pattern, with embodiment as a negative predictor and identification as a positive predictor of performance. These results open avenues for further research on the processes involved in the Proteus effect, particularly those focused on the relation between the avatar and the nature of the task to be performed. All supplemental materials are available at https://osf.io/au3wk/.
- Open Access
- 24 November 2024
- Published by Elsevier in Journal of Creativity
- Vol. 35 (1)
Multisensory objects’ role on creativity
- 24 November 2024
- Published by Elsevier in Journal of Creativity
- Vol. 35 (1)
Open AccessMultisensory objects’ role on creativity