[2025 – 2026] Building an open synthetic benchmark and reference baselines for V2X cooperative 3D semantic occupancy prediction, so that connected vehicles and roadside infrastructure can jointly reconstruct dense, semantically rich representations of the driving scene.Read more

