Hippolyte Fournier
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Hippolyte Fournier has a total of 36 co-authors. Here are the top co-authors by number of shared publications:
- Olivier Koenig
EMC, Universit Lyon 2, Lyon, France
9 shared publications - Fabien Ringeval
Université Grenoble Alpes, Grenoble, France
5 shared publications - Gérard Bailly
Grenoble Images Parole Signal Automatique, UMR 5216, University Grenoble Alpes, CNRS, Grenoble INP, France
4 shared publications - Béatrice Bouchot
ATOS Company, Bezons, France
3 shared publications - Eléonore Trân
EMC, Universit Lyon 2, Lyon, France
3 shared publications - Franck Tarpin-Bernard
Scientific Brain Training (SBT) Humans Matter, Lyon, France
3 shared publications
Latest Publications
- Open Access
- 2 November 2026
- Published by Center for Open Science in PsyArXiv
A Relevance Rate Model of the affective mechanisms linking social media use to later attentional disruption
Social media use is increasingly suspected of harming attention, but the evidence remains too limited to draw any firm conclusions. One obstacle is the lack of a framework specifying how such use might affect attention afterwards, without any device. To address this gap, we draw on emotional attention, the affective mechanism that prioritises what matters to individuals. The Relevance Rate Model rests on the relevance rate, the frequency with which an environment delivers concern-relevant stimuli. Drawing on predictive coding, we suggest that social media raises the rate that users expect, and that this expectation may persist after each session and, with repeated use, across daily life. When an activity delivers less concern-relevant information than expected, sustained attention to the ongoing task may be disrupted as emotional attention is recruited to address the shortfall. The model specifies four measurable parameters of use, from which we derive a research agenda.
- 2 November 2026
- Published by Center for Open Science in PsyArXiv
Open AccessA Relevance Rate Model of the affective mechanisms linking social media use to later attentional disruption
Social media use is increasingly suspected of harming attention, but the evidence remains too limited to draw any firm conclusions. One obstacle is the lack of a framework specifying how such use might affect attention afterwards, without any device. To address this gap, we draw on emotional attention, the affective mechanism that prioritises what matters to individuals. The Relevance Rate Model rests on the relevance rate, the frequency with which an environment delivers concern-relevant stimuli. Drawing on predictive coding, we suggest that social media raises the rate that users expect, and that this expectation may persist after each session and, with repeated use, across daily life. When an activity delivers less concern-relevant information than expected, sustained attention to the ongoing task may be disrupted as emotional attention is recruited to address the shortfall. The model specifies four measurable parameters of use, from which we derive a research agenda.
- Open Access
- 21 July 2026
- Published by Center for Open Science in PsyArXiv
Anticipatory Threat Disrupts Attention and Working Memory as a Function of Cognitive Anxiety: A Replication and Extension
Whereas the impact of threat anticipation on attentional resources has been documented, its extension to working memory (WM) had yet to be directly examined. A recent study provided first evidence that anticipating threat impairs WM, and that this impairment increased with cognitive anxiety. However, this study used ambient music inductions prior to each trial, leaving open whether observed costs reflect anticipation per se or music induction. The present study replicated this paradigm using a gamified complex-span task in which participants navigated the rooms of a virtual mansion where a neutral host and a threatening host were said to potentially appear at any time. Critically, no host ever appeared during the task. Music inductions were removed to isolate purely anticipatory processes. The aim was to test whether anticipating threat impaired attentional control and WM, and whether these effects increased with cognitive anxiety. Attentional capture and WM were indexed by parity-judgment and recall accuracy, respectively. A significant interaction with cognitive anxiety emerged, so that threat anticipation impaired both attention and WM at high anxiety. These findings replicate and refine the evidence that anticipatory threat impairs attention and WM depends critically on threat intensity and individual differences in anxiety. Moreover, it introduces a new paradigm for isolating the cognitive costs of threat anticipation.
- 21 July 2026
- Published by Center for Open Science in PsyArXiv
Open AccessAnticipatory Threat Disrupts Attention and Working Memory as a Function of Cognitive Anxiety: A Replication and Extension
Whereas the impact of threat anticipation on attentional resources has been documented, its extension to working memory (WM) had yet to be directly examined. A recent study provided first evidence that anticipating threat impairs WM, and that this impairment increased with cognitive anxiety. However, this study used ambient music inductions prior to each trial, leaving open whether observed costs reflect anticipation per se or music induction. The present study replicated this paradigm using a gamified complex-span task in which participants navigated the rooms of a virtual mansion where a neutral host and a threatening host were said to potentially appear at any time. Critically, no host ever appeared during the task. Music inductions were removed to isolate purely anticipatory processes. The aim was to test whether anticipating threat impaired attentional control and WM, and whether these effects increased with cognitive anxiety. Attentional capture and WM were indexed by parity-judgment and recall accuracy, respectively. A significant interaction with cognitive anxiety emerged, so that threat anticipation impaired both attention and WM at high anxiety. These findings replicate and refine the evidence that anticipatory threat impairs attention and WM depends critically on threat intensity and individual differences in anxiety. Moreover, it introduces a new paradigm for isolating the cognitive costs of threat anticipation.
- 7 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Interaction de la pertinence intrinsèque et de la pertinence au but dans l'attention émotionnelle et la tendance à l'action : Evidences expérimentales et propositions théorique et méthodologique
Selon les théories de l’appraisal, les interactions des individus avec leur environnement pourraient être définies sur la base d’un processus d’évaluation cognitive des stimuli selon plusieurs critères. Dans cette thèse, nous nous sommes intéressés à la manière dont le processus d’évaluation cognitive impacte les sous-systèmes cognitifs à travers l’attention par le biais de l’évaluation des stimuli selon le critère de pertinence. Notre intérêt s’est porté sur l’impact combiné de deux sous-critères du critère de pertinence sur l’attention, à savoir : la pertinence intrinsèque, définie par la nature plutôt déplaisante ou plaisante des stimuli de manière générale ; et la pertinence aux buts, définie par le niveau d’obstruction ou de facilitation des stimuli pour la réalisation des buts et besoins momentanément importants. Tandis que le caractère déplaisant et obstructif pour les buts et besoins des stimuli générerait probablement une tendance à les éviter, leur caractère plaisant et facilitateur générerait probablement une tendance à l’approche. Quatre expériences ont été réalisées afin d’apporter des données à cette thématique. Les résultats de ces études ont montré que : 1) la pertinence intrinsèque et la pertinence aux buts et besoins ont des effets multiplicatifs sur l’attention ; 2) la capture attentionnelle vers les stimuli est augmentée lorsque leurs pertinences génèrent à la fois approche et évitement ; 3) la capture attentionnelle vers des stimuli intrinsèquement pertinents peut être annulée en rendant ces stimuli non pertinents pour la réalisation d’un but induit auprès des participants ; 4) une mesure plus fine du biais attentionnel vers la menace, exacerbé dans l’anxiété, peut être obtenue au moyen d’une tâche que nous avons développée, dite « d’eyedot-probe », comparativement à une tâche de « dot-probe » classique. Enfin, un essai théorique a été réalisé afin d’intégrer les hypothèses des théories de l’appraisal à celles de la théorie de l’inférence active (i.e., définissant le cerveau comme une machine à prédire). Cet essai a permis de proposer un modèle computationnel de l’attention émotionnelle dérivé de ces deux théories.
- 7 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Interaction de la pertinence intrinsèque et de la pertinence au but dans l'attention émotionnelle et la tendance à l'action : Evidences expérimentales et propositions théorique et méthodologique
Selon les théories de l’appraisal, les interactions des individus avec leur environnement pourraient être définies sur la base d’un processus d’évaluation cognitive des stimuli selon plusieurs critères. Dans cette thèse, nous nous sommes intéressés à la manière dont le processus d’évaluation cognitive impacte les sous-systèmes cognitifs à travers l’attention par le biais de l’évaluation des stimuli selon le critère de pertinence. Notre intérêt s’est porté sur l’impact combiné de deux sous-critères du critère de pertinence sur l’attention, à savoir : la pertinence intrinsèque, définie par la nature plutôt déplaisante ou plaisante des stimuli de manière générale ; et la pertinence aux buts, définie par le niveau d’obstruction ou de facilitation des stimuli pour la réalisation des buts et besoins momentanément importants. Tandis que le caractère déplaisant et obstructif pour les buts et besoins des stimuli générerait probablement une tendance à les éviter, leur caractère plaisant et facilitateur générerait probablement une tendance à l’approche. Quatre expériences ont été réalisées afin d’apporter des données à cette thématique. Les résultats de ces études ont montré que : 1) la pertinence intrinsèque et la pertinence aux buts et besoins ont des effets multiplicatifs sur l’attention ; 2) la capture attentionnelle vers les stimuli est augmentée lorsque leurs pertinences génèrent à la fois approche et évitement ; 3) la capture attentionnelle vers des stimuli intrinsèquement pertinents peut être annulée en rendant ces stimuli non pertinents pour la réalisation d’un but induit auprès des participants ; 4) une mesure plus fine du biais attentionnel vers la menace, exacerbé dans l’anxiété, peut être obtenue au moyen d’une tâche que nous avons développée, dite « d’eyedot-probe », comparativement à une tâche de « dot-probe » classique. Enfin, un essai théorique a été réalisé afin d’intégrer les hypothèses des théories de l’appraisal à celles de la théorie de l’inférence active (i.e., définissant le cerveau comme une machine à prédire). Cet essai a permis de proposer un modèle computationnel de l’attention émotionnelle dérivé de ces deux théories.
- Open Access
- 9 March 2026
- Published by Springer Nature in Affective Science
- Vol. 7 (2) , 293-305
When Threat Comes from Within: First Evidence that Anticipatory Emotion Impairs Working Memory
Although anticipating threat is known to capture attentional resources, its direct impact on working memory (WM), a core cognitive function supporting everyday activities, has yet to be empirically examined. Given the close dependence of WM on attentional resources, it is therefore reasonable to hypothesize that anticipating threat disrupts WM performance. Moreover, since threat-related attentional bias increases with anxiety, WM disruption caused by anticipatory threat may likewise intensify with anxiety. In Experiment 1, we tested whether anticipating threat captures attentional resources and impairs WM, and whether these effects vary with anxiety. Experiment 2 replicated and extended this design under higher task demands. Participants performed a gamified complex-span task in which each trial occurred in a virtual room associated with the potential appearance of a proximal threat (a sudden screamer), a distal threat (loss of accumulated points), or a neutral condition. Importantly, no threat ever appeared, isolating purely anticipatory processes. Attentional capture was indexed by parity-judgment accuracy, and WM by serial recall. Across both experiments, threat anticipation produced attentional capture and impaired WM. In Experiment 1, attentional capture decreased as anxiety increased, whereas this compensatory pattern disappeared under higher task load in Experiment 2. In contrast, WM impairment increased with cognitive anxiety, particularly under distal threat. These findings provide the first empirical evidence that anticipating threat impairs WM, in line with recent models of emotion, anxiety, and WM.
- 9 March 2026
- Published by Springer Nature in Affective Science
- Vol. 7 (2) , 293-305
Open AccessWhen Threat Comes from Within: First Evidence that Anticipatory Emotion Impairs Working Memory
Although anticipating threat is known to capture attentional resources, its direct impact on working memory (WM), a core cognitive function supporting everyday activities, has yet to be empirically examined. Given the close dependence of WM on attentional resources, it is therefore reasonable to hypothesize that anticipating threat disrupts WM performance. Moreover, since threat-related attentional bias increases with anxiety, WM disruption caused by anticipatory threat may likewise intensify with anxiety. In Experiment 1, we tested whether anticipating threat captures attentional resources and impairs WM, and whether these effects vary with anxiety. Experiment 2 replicated and extended this design under higher task demands. Participants performed a gamified complex-span task in which each trial occurred in a virtual room associated with the potential appearance of a proximal threat (a sudden screamer), a distal threat (loss of accumulated points), or a neutral condition. Importantly, no threat ever appeared, isolating purely anticipatory processes. Attentional capture was indexed by parity-judgment accuracy, and WM by serial recall. Across both experiments, threat anticipation produced attentional capture and impaired WM. In Experiment 1, attentional capture decreased as anxiety increased, whereas this compensatory pattern disappeared under higher task load in Experiment 2. In contrast, WM impairment increased with cognitive anxiety, particularly under distal threat. These findings provide the first empirical evidence that anticipating threat impairs WM, in line with recent models of emotion, anxiety, and WM.
- Top Cited PapersOpen Access
- 29 January 2026
- Published by Elsevier in Computers in Human Behavior
- Vol. 179
Attention hijacked: How social media notifications disrupt cognitive processing
- 29 January 2026
- Published by Elsevier in Computers in Human Behavior
- Vol. 179
Top Cited PapersOpen AccessAttention hijacked: How social media notifications disrupt cognitive processing
- Open Access
- 30 September 2025
- Published by Center for Open Science in PsyArXiv
Attention hijacked: How social media notifications disrupt cognitive processing
Digital notifications are ubiquitous in modern life, yet their impact on cognitive performance remains poorly understood. Here, we introduce an ecologically valid paradigm to examine how social media notifications disrupt ongoing task performance. Participants completed a Stroop task while receiving smartphone-style notifications, allowing us to track both behavioural and physiological responses. Notifications triggered a transient slowdown in cognitive processing lasting approximately seven seconds, driven by the combined influence of perceptual salience, learned associations, and relevance appraisal. The magnitude of this disruption was predicted by the inferred relevance of notifications and by individual patterns of smartphone use—specifically, the frequency of interaction (notification volume and checking behaviour) rather than total time spent on the device. These findings were mirrored in changes in pupil dilation, a physiological marker of cognitive load, confirming the convergent impact of notification relevance and smartphone habits. Together, our results show that modern digital cues can hijack attentional resources, even in the absence of explicit personal relevance. They underscore the need to account for notification frequency—not merely screen time—when evaluating the cognitive cost of pervasive digital environments, and raise broader concerns about the long-term impact of habitual device interaction on attentional functioning.
- 30 September 2025
- Published by Center for Open Science in PsyArXiv
Open AccessAttention hijacked: How social media notifications disrupt cognitive processing
Digital notifications are ubiquitous in modern life, yet their impact on cognitive performance remains poorly understood. Here, we introduce an ecologically valid paradigm to examine how social media notifications disrupt ongoing task performance. Participants completed a Stroop task while receiving smartphone-style notifications, allowing us to track both behavioural and physiological responses. Notifications triggered a transient slowdown in cognitive processing lasting approximately seven seconds, driven by the combined influence of perceptual salience, learned associations, and relevance appraisal. The magnitude of this disruption was predicted by the inferred relevance of notifications and by individual patterns of smartphone use—specifically, the frequency of interaction (notification volume and checking behaviour) rather than total time spent on the device. These findings were mirrored in changes in pupil dilation, a physiological marker of cognitive load, confirming the convergent impact of notification relevance and smartphone habits. Together, our results show that modern digital cues can hijack attentional resources, even in the absence of explicit personal relevance. They underscore the need to account for notification frequency—not merely screen time—when evaluating the cognitive cost of pervasive digital environments, and raise broader concerns about the long-term impact of habitual device interaction on attentional functioning.
- Open Access
- 23 July 2025
- Published by Center for Open Science in PsyArXiv
When threat comes from within: first evidence that internal appraisal impairs working memory
Emotional stimuli, especially threatening ones, are known to consume attentional resources, and thus impair working memory (WM). However, this effect has only been demonstrated with perceptual threats (e.g., images of snakes), whereas real-life threats often arise from internal cognitive appraisal of possible negative outcomes. Although this has never been directly tested, WM models suggest that internal appraisals should also impair WM. We tested this hypothesis in two experiments using a novel paradigm that induced internal threat appraisals, under low and high task demands, during a complex span task. Across both experiments, threat appraisals impaired WM performance. Crucially, under high task demand, this impairment was amplified in individuals with elevated cognitive anxiety, particularly when the threat required greater cognitive elaboration. This study introduces a novel paradigm to assess internally generated threat appraisal and demonstrates, for the first time, that it disrupts WM processes similarly to external emotional stimuli.
- 23 July 2025
- Published by Center for Open Science in PsyArXiv
Open AccessWhen threat comes from within: first evidence that internal appraisal impairs working memory
Emotional stimuli, especially threatening ones, are known to consume attentional resources, and thus impair working memory (WM). However, this effect has only been demonstrated with perceptual threats (e.g., images of snakes), whereas real-life threats often arise from internal cognitive appraisal of possible negative outcomes. Although this has never been directly tested, WM models suggest that internal appraisals should also impair WM. We tested this hypothesis in two experiments using a novel paradigm that induced internal threat appraisals, under low and high task demands, during a complex span task. Across both experiments, threat appraisals impaired WM performance. Crucially, under high task demand, this impairment was amplified in individuals with elevated cognitive anxiety, particularly when the threat required greater cognitive elaboration. This study introduces a novel paradigm to assess internally generated threat appraisal and demonstrates, for the first time, that it disrupts WM processes similarly to external emotional stimuli.
- 3 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Affective Computing
- Vol. 16 (3) , 2233-2244
THERADIA WoZ: An Ecological Corpus for Appraisal-Based Affect Research in Healthcare
We present THERADIA WoZ, an ecological corpus designed for audiovisual research on affect in healthcare. Two groups of senior individuals, consisting of 52 healthy participants and 9 individuals with Mild Cognitive Impairment (MCI), performed Computerised Cognitive Training (CCT) exercises while receiving support from a virtual assistant, tele-operated by a human in the role of a Wizard-of-Oz (WoZ). The audiovisual expressions produced by the participants were fully transcribed, and partially annotated based on dimensions derived from recent appraisal theory models, including novelty, intrinsic pleasantness, goal conduciveness, and coping. Additionally, the annotations included 23 affective labels from the literature of achievement affects. We present the data collection, transcription, and annotation protocols, alongside a detailed analysis of the annotated dimensions and labels. Baseline methods and results for their automatic prediction are also presented. Results reveal that the dimensions of appraisal theory can be predicted, with the performance varying across different modalities. The corpus aims to serve as a valuable resource for researchers in affective computing, and is made available to both industry and academia.
- 3 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Affective Computing
- Vol. 16 (3) , 2233-2244
THERADIA WoZ: An Ecological Corpus for Appraisal-Based Affect Research in Healthcare
We present THERADIA WoZ, an ecological corpus designed for audiovisual research on affect in healthcare. Two groups of senior individuals, consisting of 52 healthy participants and 9 individuals with Mild Cognitive Impairment (MCI), performed Computerised Cognitive Training (CCT) exercises while receiving support from a virtual assistant, tele-operated by a human in the role of a Wizard-of-Oz (WoZ). The audiovisual expressions produced by the participants were fully transcribed, and partially annotated based on dimensions derived from recent appraisal theory models, including novelty, intrinsic pleasantness, goal conduciveness, and coping. Additionally, the annotations included 23 affective labels from the literature of achievement affects. We present the data collection, transcription, and annotation protocols, alongside a detailed analysis of the annotated dimensions and labels. Baseline methods and results for their automatic prediction are also presented. Results reveal that the dimensions of appraisal theory can be predicted, with the performance varying across different modalities. The corpus aims to serve as a valuable resource for researchers in affective computing, and is made available to both industry and academia.
- Open Access
- 11 March 2025
- Published by Center for Open Science in PsyArXiv
Evidence of an exponential relationship between attentional bias to threat and anxiety using an original “eye-dot-probe” task
Over the past decade, the anxiety-related attentional bias to threat, widely investigatedthrough the dot-probe task, has faced challenges. First, at an early stage of processing, itremains unclear whether this bias follows a static pattern (vigilance) or fluctuates betweenvigilance and avoidance. Second, methodological and theoretical concerns have been raisedregarding the reliability of the dot-probe task in measuring threat-related attentional bias. Thepresent study aimed to test the hypothesis of a static early expression of attentional bias tothreat using both a classical dot-probe task and a novel eye-dot-probe task, in which responseswere made via ocular saccades. Additionally, we hypothesized that the relationship betweenattentional bias to threat and anxiety follows an exponential rather than linear trajectory. Non-clinical participants completed both tasks along with an anxiety questionnaire. Resultsshowed no consistent relationship between anxiety and either static or fluctuating attentionalbias in the classical dot-probe task. However, in the eye-dot-probe task, an exponentialrelationship emerged between anxiety and static attentional bias toward threat, with asignificant bias observed in high-anxious individuals. These findings provide theoretical andmethodological insights, suggesting that early attentional bias toward threat is static, bestcaptured using eye-movement responses, and follows a non-linear relationship with anxiety,emerging only beyond a critical threshold.
- 11 March 2025
- Published by Center for Open Science in PsyArXiv
Open AccessEvidence of an exponential relationship between attentional bias to threat and anxiety using an original “eye-dot-probe” task
Over the past decade, the anxiety-related attentional bias to threat, widely investigatedthrough the dot-probe task, has faced challenges. First, at an early stage of processing, itremains unclear whether this bias follows a static pattern (vigilance) or fluctuates betweenvigilance and avoidance. Second, methodological and theoretical concerns have been raisedregarding the reliability of the dot-probe task in measuring threat-related attentional bias. Thepresent study aimed to test the hypothesis of a static early expression of attentional bias tothreat using both a classical dot-probe task and a novel eye-dot-probe task, in which responseswere made via ocular saccades. Additionally, we hypothesized that the relationship betweenattentional bias to threat and anxiety follows an exponential rather than linear trajectory. Non-clinical participants completed both tasks along with an anxiety questionnaire. Resultsshowed no consistent relationship between anxiety and either static or fluctuating attentionalbias in the classical dot-probe task. However, in the eye-dot-probe task, an exponentialrelationship emerged between anxiety and static attentional bias toward threat, with asignificant bias observed in high-anxious individuals. These findings provide theoretical andmethodological insights, suggesting that early attentional bias toward threat is static, bestcaptured using eye-movement responses, and follows a non-linear relationship with anxiety,emerging only beyond a critical threshold.
- Open Access
- 4 November 2024
- Published by Association for Computing Machinery (ACM)
- p. 677-683
EVAC 2024 – Empathic Virtual Agent Challenge: Appraisal-based Recognition of Affective States
As autonomous interactive agents become increasingly prevalent, it is crucial for these virtual agents to understand and respond to both our verbal content and emotions, enabling deeper interactions. Despite significant advancements in the automatic recognition and understanding of human speech, challenges remain in accurately identifying and addressing the nuances of human emotions, hindering the development of more empathic virtual agents. We believe that empathic virtual agents should excel in three key tasks: (i) recognising spontaneous emotional expressions alongside understanding verbal content, (ii) generating timely and appropriate responses, and (iii) providing insightful feedback while comprehending user responses. To advance the development of empathic agents, we introduce the first Empathic Virtual Agent Challenge (EVAC). The inaugural edition focuses on robustly recognising spontaneous human expressions during interactions with a virtual agent, using the newly introduced THERADIA WoZ dataset. This paper provides an overview of the baseline systems operated on the pseudonymised version of the corpus on the two following modeling tasks: core affect presence and intensity, and appraisal based dimensions.
- 4 November 2024
- Published by Association for Computing Machinery (ACM)
- p. 677-683
Open AccessEVAC 2024 – Empathic Virtual Agent Challenge: Appraisal-based Recognition of Affective States
As autonomous interactive agents become increasingly prevalent, it is crucial for these virtual agents to understand and respond to both our verbal content and emotions, enabling deeper interactions. Despite significant advancements in the automatic recognition and understanding of human speech, challenges remain in accurately identifying and addressing the nuances of human emotions, hindering the development of more empathic virtual agents. We believe that empathic virtual agents should excel in three key tasks: (i) recognising spontaneous emotional expressions alongside understanding verbal content, (ii) generating timely and appropriate responses, and (iii) providing insightful feedback while comprehending user responses. To advance the development of empathic agents, we introduce the first Empathic Virtual Agent Challenge (EVAC). The inaugural edition focuses on robustly recognising spontaneous human expressions during interactions with a virtual agent, using the newly introduced THERADIA WoZ dataset. This paper provides an overview of the baseline systems operated on the pseudonymised version of the corpus on the two following modeling tasks: core affect presence and intensity, and appraisal based dimensions.