WHU-Infra3D: A Full-stack Multi-modal Dataset and Benchmark for 3D Roadside Infrastructure Inventory
arXiv:2606.09882v1 Announce Type: cross Abstract: The paradigm of digital twin cities is shifting from coarse visual mapping toward more precise and actionable digitization of urban assets. However, existing datasets predominantly focus on coarse visual perception, lacking the strict multi-modal alignment and attribute and status diagnosis required for automated infrastructure maintenance. To bridge this gap, we introduce WHU-Infra3D, a large-scale, multi-modal benchmark dataset dedicated to roadside infrastructure inventory. Covering 53.8 km across three cities, WHU-Infra3D uniquely integrate
The increasing sophistication of digital twin initiatives and the growing demand for automated infrastructure maintenance are driving the need for more precise and actionable urban asset digitization.
This new dataset provides a critical foundation for developing advanced AI models capable of detailed, multi-modal analysis of urban infrastructure, accelerating the move towards automated and predictive maintenance.
The availability of WHU-Infra3D shifts the focus from coarse visual mapping to highly precise, multi-modal data integration for infrastructure inventory, significantly enhancing the potential for AI-driven urban asset management.
- · AI developers in urban planning
- · Smart city technology providers
- · Infrastructure maintenance companies
- · Sensors and data collection hardware manufacturers
- · Manual infrastructure inspection services
- · Legacy GIS and mapping solutions
- · Data companies focused solely on visual-only modalities
Improved efficiency and accuracy in urban infrastructure monitoring and maintenance through AI.
Reduced operational costs and extended lifespans of urban assets due to proactive and data-driven interventions.
Enhanced urban resilience and safety, enabling cities to better anticipate and respond to infrastructure failures.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG