This research examines how architectural and environmental features shape emotional experience in adolescents with emotional dysregulation and social withdrawal, framing representation as an active methodological tool. Within the Italian tradition of Representation, images and virtual environments enable the isolation and manipulation of spatial variables that are otherwise difficult to quantify, providing controlled yet ecologically valid stimuli. The EmoVirtual protocol includes two studies: EVIT, where 60 systematically varied images are rated by three groups (controls, DE, HI) on Preference, Valence, and Arousal and analyzed through multi-layer computational features; and EVIE, which translates these variables into high-fidelity VR environments (Gaussian Splatting) to assess effects on anxiety, affect, and presence. Results show a consistent preference for natural settings, while built environments are associated with less adaptive responses. Clinical groups exhibit distinct patterns: high variability (DE) and generalized hypoactivation (HI). VR exposure suggests preliminary regulatory benefits, including anxiety reduction and affective improvement. The study demonstrates how representation-driven tools can be integrated into clinical research and supports a shift toward environments designed for psychological as well as physical inclusion.

EmoVirtual: Environmental Representation and Virtual Reality as Conceptual and Operational Frameworks for Investigating Adolescent Affective Response and Regulation in Clinical Contexts

RECHICHI, PIERGIUSEPPE
2026

Abstract

This research examines how architectural and environmental features shape emotional experience in adolescents with emotional dysregulation and social withdrawal, framing representation as an active methodological tool. Within the Italian tradition of Representation, images and virtual environments enable the isolation and manipulation of spatial variables that are otherwise difficult to quantify, providing controlled yet ecologically valid stimuli. The EmoVirtual protocol includes two studies: EVIT, where 60 systematically varied images are rated by three groups (controls, DE, HI) on Preference, Valence, and Arousal and analyzed through multi-layer computational features; and EVIE, which translates these variables into high-fidelity VR environments (Gaussian Splatting) to assess effects on anxiety, affect, and presence. Results show a consistent preference for natural settings, while built environments are associated with less adaptive responses. Clinical groups exhibit distinct patterns: high variability (DE) and generalized hypoactivation (HI). VR exposure suggests preliminary regulatory benefits, including anxiety reduction and affective improvement. The study demonstrates how representation-driven tools can be integrated into clinical research and supports a shift toward environments designed for psychological as well as physical inclusion.
12-lug-2026
Inglese
affective response
computer vision
environmental psychology
gaussian splatting
image feature analysis
virtual reality
Bevilacqua, Marco Giorgio
Milone, Annarita
Piemonte, Andrea
File in questo prodotto:
File Dimensione Formato  
01_Rechichi_Tesi_pdfA.pdf

embargo fino al 14/07/2029

Licenza: Creative Commons
Dimensione 121.42 MB
Formato Adobe PDF
121.42 MB Adobe PDF

I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.14242/384946
Il codice NBN di questa tesi è URN:NBN:IT:UNIPI-384946