Introduction
The UNESCO Virtual Museum of Stolen Cultural Objects was created to make absence visible. Each stolen artifact represents a fragment of cultural memory removed from its community. Our mission was to translate that absence into a digital experience that feels both immersive and respectful, where technology serves memory rather than spectacle. The project took nearly two years of pitching and alignment, followed by almost a year of design and development with a multidisciplinary team of more than 25 people, including art directors, UX designers, 3D artists, creative developers, strategists, and producers. Rather than recreating a traditional museum online, we built a living digital space inspired by architect Francis Kéré’s vision of an open, fluid environment where visitors explore freely.
Reconstructing the Invisible
One of the most complex parts of the project was reconstructing stolen objects using incomplete documentation, often sourced from the Interpol database. Some objects were represented by only one low-quality photograph. We developed a hybrid workflow combining AI reconstruction and manual 3D craftsmanship. AI could generate surprisingly accurate base meshes for some materials, especially stone artifacts, but often struggled with precision, perspective, or texture fidelity.
The “Caliz 3” is a good example. AI only reconstructed the upper section correctly. Aurélien, our 3D artist, manually rebuilt the missing geometry and merged it with the usable AI output.
The “Ring” presented another challenge. Perspective inconsistencies and strong contrast in the source image produced unusable textures with large black areas. Objects containing gems or crystals were particularly difficult because real-time rendering constraints prevented physically accurate refraction.
The “Altar sculpture of Virgin and Child with Saint Anne” introduced an unexpected obstacle: AI moderation systems refused to process the images due to nudity. The team created a workaround by retouching and recoloring the source images before reconstructing missing elements manually.
Not every moment was friction. The “Zoomorphic Figure”, loved by Aurélien, became a small symbol of joy in the process, an odd, charming creature somewhere between a lamb and a frog that reminded us why the objects themselves mattered.
“AI accelerated the process, but craftsmanship made the objects believable.”
A “procedural” generative Baobab
At the center of the museum’s spatial concept lies a generative “baobab” structure inspired by Francis Kéré’s vision of cultural objects as the roots of communities. The museum grows digitally like a living structure. Instead of fixed galleries, the space unfolds organically, branching into new perspectives as visitors explore.
To maintain performance, galleries and 3D objects are progressively streamed and instantiated in the scene based on the visitor’s scroll position, while distant elements are unloaded or simplified through level-of-detail management. This ensures the experience remains fluid even as the environment continuously expands.
This infinite-scroll exploration was intentionally designed to echo how younger generations discover content today on social platforms: through fluid, continuous movement rather than hierarchical navigation. By borrowing this interaction logic and translating it into a spatial experience, the museum becomes immediately intuitive to navigate, while still preserving a sense of wonder and discovery. The generative baobab is therefore both a symbolic structure and a navigation system, connecting memory, architecture, and contemporary digital behavior into a single experience.
Designing a Museum Without Walls
The museum navigation reflects Francis Kéré’s architectural philosophy: openness, fluidity, and continuity. Instead of rooms connected by corridors, visitors scroll through evolving spaces that feel limitless.
Jerome, Romain, and Solène shaped the artistic direction of the environments, translating architectural intentions into digital materials, lighting, and spatial rhythm, with the spaces unified and enhanced by symbolic light wells reminiscent of constellations.
Performance was a constant constraint. Every object had to feel detailed and tangible while remaining lightweight enough to load smoothly on the web. Creative developers Gauthier, Nils, and Florentin worked closely with the creative team to optimize models and textures without losing visual fidelity.
Hélène, the producer, played a crucial role in maintaining alignment between UNESCO’s cultural mission and the technical realities of real-time 3D production.
Crafting the Interface: Where 3D Meets Editorial Design
Beyond the architecture and reconstruction work, an enormous amount of care went into crafting the interface itself. The museum had to feel immersive, but also readable, accessible, and editorial. Jerome, Romain, and Solène shaped a visual language that bridges spatial 3D environments with refined 2D typography and layout. Every transition, from object focus to information page, was designed to feel intentional and fluid, never abrupt.
Our front-end team, Vincent, Zoé, and Rémi, worked closely with the design team to polish micro-interactions and animation timing. The interface doesn’t sit on top of the 3D world; it emerges from it. When a visitor selects an object, the camera subtly reframes, the background softens, and the experience transitions seamlessly from real-time 3D space to a crafted 2D editorial layout. This choreography ensures continuity between immersion and information.
Each cultural object opens into a dedicated page that expands its story. These pages are not simple overlays; they are carefully composed spaces featuring contextual texts, historical insights, and, crucially, the original Interpol photograph used to reconstruct the object. Showing the source image was an ethical and design decision: it reinforces transparency and reminds visitors of the fragmentary documentation behind each reconstruction.
The museum also empowers visitors to become curators. Through an advanced filtering system, by region, category, date, usage, technique, material, or even dominant color, users can generate their own personalized galleries. These dynamic collections are assembled in real time, transforming the museum from a static archive into a participatory space. This feature was especially important for engaging younger audiences, who are used to discovering and organizing content based on personal interest rather than institutional hierarchy.
WebVR and Real-Time Rendering with NanoGL
From the beginning, we wanted the museum to exist beyond the screen, not just as a website, but as a place visitors could inhabit. That ambition led us to develop a full WebVR version of the experience using WebXR, powered by NanoGL, our proprietary WebGL engine.
NanoGL became the backbone of the real-time experience. It allowed us to render complex architectural environments and dozens of reconstructed cultural objects directly in the browser while maintaining stable performance across devices. Because VR demands strict frame-rate stability, the engine required significant adaptation: scene streaming, progressive loading, and aggressive geometry optimization were essential to keep the experience smooth without compromising visual fidelity.
In VR, interaction design shifts dramatically. Navigation is no longer about scrolling; it becomes spatial and physical. Visitors can move freely through the museum, approach objects, and examine them at human scale. This change transforms perception: a small pendant suddenly feels intimate and fragile, while larger sculptures gain presence and weight.
More than a technical extension, the VR version became an emotional amplifier of the project’s mission. When visitors stand inside the museum space and encounter reconstructed objects at scale, the sense of absence becomes tangible. The experience shifts from observation to presence, from browsing a website to visiting a place of memory.
Where Technology Meets Craft
Over the course of the project, the museum slowly transformed from a technical challenge into something more personal. Each reconstructed object reminded us that behind every missing artifact lies a story interrupted, a community, a ritual, a fragment of identity displaced.
Technology made reconstruction possible, but it was human care that gave the objects presence again. AI could suggest shapes, generate textures, and accelerate workflows, but it was the craft and skill of designers and developers that turned imperfect references into meaningful representations. Every manual correction, every lighting adjustment, every optimization decision became part of a collective effort to treat these objects with respect.
Production unfolded as a rhythm of experimentation and discovery. Some reconstructions came together quickly, almost effortlessly. Others required days of iteration, uncertainty, and persistence. Those slower moments often became the most meaningful ones, when the team moved from solving technical problems to understanding the cultural weight of what they were rebuilding.
The VR experience made this especially tangible. Standing inside the museum, surrounded by reconstructed objects, the absence they represent becomes impossible to ignore. The digital space stops feeling like an interface and begins to feel like a place of memory.
By the end of the journey, the project felt less like building a website and more like assembling fragments of history into a shared space. The UNESCO Virtual Museum of Stolen Cultural Objects does not attempt to replace what was stolen. Instead, it acknowledges loss while keeping memory visible and accessible.
If the project proves anything, it is that technology can be more than a tool for innovation: it can be a medium for remembrance, empathy, and connection across cultures and time.
Technologies
Frontend Frameworks and Libraries: Nuxt.js, Tailwind, Typescript, WebGL, NanoGL
Backend Technologies: Node.js
Server Architecture: Cloud-hosted content delivery optimized for 3D assets
Tools: Blender, Cinema 4D, AI generative reconstruction pipelines, WebXR
Company Info
makemepulse is a digital creative studio exploring the intersection of storytelling, design, and technology. With teams spanning creative direction, real-time development, 3D production, and UX design, the studio builds digital experiences that transform complex cultural and technological topics into meaningful interactive narratives.