Neural Artifact
Lab

The Neural Artifact Lab develops AI systems that do not simply process visual data, but seek to understand the relationships, contexts, and material realities that shape it. We believe that visual intelligence emerges through structure, interaction, and interpretation: from how objects occupy space, to how fragments relate to one another, to how meaning is constructed through cultural and human context.

Our research combines computer vision, machine learning, spatial computing, and multimodal AI with perspectives drawn from the arts, humanities, and cultural heritage. Rather than treating these domains as applications alone, we see them as critical environments for developing more interpretable, context-aware, and human-centred forms of artificial intelligence.

The lab’s work explores how AI can support reasoning about incomplete, ambiguous, or evolving worlds — reconstructing damaged artefacts, understanding complex scenes, linking visual and textual knowledge, and enabling new forms of interaction with cultural and material data. Across these areas, we aim to build systems that augment human expertise, encourage exploration and interpretation, and foreground uncertainty, context, and critical engagement as central components of intelligent technologies.

We collaborate extensively with partners across Europe through EU Horizon projects and engage closely with the Durham Centre for Digital Humanities.