EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Identification of whole and partial facial expressions of human multimodal emotional states

V.A. Barabanschikov, E.V. Suvorova, A.S. Borisova

Experimental Psychology (Russia) · 30.03.2026 · 19 (1) · 4-21

DOI: 10.17759/exppsy.2026190101 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Experimental Psychology (Russia), online erschienen am 30.03.2026. Verfasst von V.A. Barabanschikov und 2 weiteren Autor:innen. Die Arbeit nennt 41 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Context and relevance. Within the framework of the cognitive-communicative approach, the mechanisms of perception of multimodal expressions of human affective states are studied. The article considers the specifics of perception of bimodal emotional states of a person when exposing a whole and partially open face. Objective: comparative analysis of the productivity of perception of bimodal (face + voice) expressions of emotions in conditions of a fully and partially open face. Methods and materials. The experiment is based on the Russian version of the Geneva Emotion Recognition Test using facial occlusions. It consists of three episodes with 40 participants each; ages are 18—52 years old. The subjects were consistently presented with short (3—5 s) audio video clips of key affective states expressed by specially trained professional actors on the monitor screen. In the first series, their faces were completely exposed; in the second and third, the upper or lower halves of their faces were covered with a rectangular mask. It was necessary to determine the emotional state of the sitters by touching the corresponding button on the interactive scale of the Geneva Emotion Wheel.

Results

New data has been obtained regarding the correctness of recognition of bimodal expressions when exposing a whole and partially open face, their relationship and dependence on the content of emotions and the location of occlusion. It is shown that when perceiving bimodal expressions, the popular maxim “the whole is greater than the part” is limited in nature, not taking into account the specifics of each of the unimodal expressions and their interrelationships. The content of emotions and the area of facial occlusion play an essential role. The types of multimodal relations of a part and a whole were revealed, covering different groups of affective categories: “the whole is greater than each of the parts” (36% of states), “the whole is equal to each of the parts” (21%), “the whole is greater than one of the parts” (14%) and “the whole is less than one of the parts” (28%). In 57% of cases, the accuracy of identification of the expressions of the upper and lower halves of the same face is indistinguishable, with the divergence of values, the expressions of the upper half are more accurately perceived.

Conclusions

In conditions of multimodality, each of the halves of the face has its own expressive resources that are more closely linked to the potential of the face as a whole. The “paradoxical” effects of perceiving the expression of the whole and part of the face are due to the process of cross-modal integration of facial and prosodic forms of bimodal states.

Zitieren

APA

Barabanschikov, V., Suvorova, E., & Borisova, A. (2026). Identification of whole and partial facial expressions of human multimodal emotional states. Experimental Psychology (Russia), 19(1), 4-21. https://doi.org/10.17759/exppsy.2026190101

Vancouver

Barabanschikov V, Suvorova E, Borisova A. Identification of whole and partial facial expressions of human multimodal emotional states. Experimental Psychology (Russia). 2026;19(1):4-21. doi:10.17759/exppsy.2026190101

BibTeX
@article{Barabanschikov2026,
  author = {Barabanschikov, V.A. and Suvorova, E.V. and Borisova, A.S.},
  title = {Identification of whole and partial facial expressions of human multimodal emotional states},
  journal = {Experimental Psychology (Russia)},
  year = {2026},
  volume = {19},
  number = {1},
  pages = {4--21},
  doi = {10.17759/exppsy.2026190101}
}
Kostenlos starten →