From Context to Human: A Review of VLM Contextualization in the Recognition of Human States in Visual Data

This paper presents a narrative review of the contextualization and contribution offered by vision–language models (VLMs) for human-centric understanding in images. Starting from the correlation between humans and their context (background) and by incorporating VLM-generated embeddings into recognit...

Descripció completa

Dades bibliogràfiques

Autors principals:	Corneliu Florea, Constantin-Bogdan Popescu, Andrei Racovițeanu, Andreea Nițu, Laura Florea
Format:	Artigo
Idioma:	Inglês
Publicat:	MDPI AG 2026-01-01
Col·lecció:	Mathematics
Matèries:	vision–language models (VLMs) contextual bias image contextualization multimodal learning affective computing
Accés en línia:	https://www.mdpi.com/2227-7390/14/1/175
Etiquetes:	Afegir etiqueta Sense etiquetes, Sigues el primer a etiquetar aquest registre!