School Legacy: Reconstructing Memory Through AI-augmented Social XR

Yangli Liu, Mehdi Akhondi, Eraulia Nadinda Rachmat, Mengyao Guo, Ibrahim Olalekan Jimoh, and Haowei Xiong. 2025. School Legacy: Reconstructing Memory Through AI-augmented Social XR. In SIGGRAPH Asia 2025 Posters (SA Posters ’25), December 15-18, 2025. ACM, New York, NY, USA, 3 pages. (CCF-A)

Objective:
Address the limitations of traditional memory preservation, particularly the lack of spatial, social, and emotional reconstruction, by enabling shared and embodied memory experiences through AI-augmented social XR.

Methods:

· Designed and developed School Legacy, an XR system integrating MR and VR to reconstruct personal and collective memories.

· Built an AI-driven pipeline to transform personal 2D images into textured 3D “time capsules” for interactive use.

· Designed a full user experience flow from physical interaction (QR-based spatial anchoring) to multiplayer VR social engagement.

· Conducted preliminary user studies to evaluate emotional engagement, usability, and social presence.

Results:

· Developed School Legacy, an AI-augmented social XR system that transforms individual memories into shared, interactive experiences.

· Demonstrated strong emotional engagement and social presence, with most participants reporting meaningful nostalgic experiences.

· The work was accepted at SIGGRAPH Asia 2025 Posters (CCF-A), validating its contribution to XR and digital heritage research.

Contribution:
Led AI-3D module development and system integration, contributing to the design of the AI generation pipeline, interactive experience flow, and overall technical implementation.

Abstract:

Traditional approaches to memory preservation, such as photographs, often fail to capture the spatial, social, and emotional dimensions of lived experiences. This limitation is especially evident in contexts such as school life, where memories are inherently collective and place-based.

To address this gap, this study presents School Legacy, an AI-augmented social XR system that transforms personal memories into shared, immersive experiences. The system integrates tangible interaction, generative AI, and multiplayer virtual environments. Users scan physical artifacts to reconstruct spatial environments, convert personal images into 3D “time capsules” using AI, and engage in synchronized social activities within a shared virtual classroom.

The system enables users to collaboratively reconstruct memory spaces and participate in nostalgic rituals such as yearbook signing, object sharing, and group interaction. Preliminary user studies indicate strong emotional engagement and enhanced social presence, highlighting the potential of XR and AI technologies in rethinking memory as a collective, embodied experience.

These findings suggest new directions for digital heritage, social XR, and memory reconstruction systems.

Keywords:
Extended Reality (XR), Generative AI, Social VR, Memory Reconstruction, Digital Heritage