Image-to-3D for Cultural Heritage: Voice Interaction and Retrieval-Augmented Guidance in VR

View at publisher Summarise with VIRAT
Authors: Rone C. Brandão Filho, Murilo De Oliveira Guimarães, Arthur Galdino Dangoni, Arthur Ricardo Sousa Vitória, Rafael Teixeira Sousa, Arlindo Rodrigues Galvão Filho
Cited by: 0
References: 28

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Virtual museum visits have an ever-growing collection due to the high demand for the digitization of historically, artistically, and culturally significant artifacts. This demand is growing even more with the popularization of Virtual Reality (VR) technology, which requires immersive and interactive experiences to motivate users to move from mere spectators to active participants. In this context, this work proposes an end-to-end approach to create a custom and interactive VR experiences through virtual exhibitions. Using a single image, artifacts from real museum exhibitions are generated and assembled into a virtual museum, enabling customized and interactive VR experiences. In addition, the museum is voice-controlled, allowing users to manipulate its display in real time, and a retrieval-augmented generation technique is employed to provide grounded, object-scoped explanations. In an initial analysis, an evaluation was performed using the LLM-as-judge protocol to analyze the quality of the reconstructions. Overall, our study demonstrated promising results in making virtual tours even more immersive and educational, creating an interactive environment and enriching the experience.

Fields of Study & Subjects

Image Processing and 3D Reconstruction 3D Surveying and Cultural Heritage Virtual Reality Applications and Impacts

Publication details

Journal / source
IEEE International Conference on Artificial Intelligence and eXtended and Virtual Reality (AIxVR)
Pages
300–304
Publication date
26 January 2026
Language
EN