IONIAN Wins 1st Place in the Mixed Signals Challenge at ITSC 2026

🏆 The IONIAN team won first place in the Mixed Signals: V2X Collaborative 3D Object Detection Challenge, held at the IEEE International Conference on Intelligent Transportation Systems (ITSC 2026)!

The challenge

The competition is built on the Mixed Signals dataset (ICCV 2025), a diverse, real-world point cloud dataset for heterogeneous LiDAR V2X collaboration. Each scene is captured by three connected vehicles and a roadside unit with dual LiDAR, giving five LiDAR streams from different viewpoints.

The goal is to detect vehicles, bikes, and pedestrians in the bird’s-eye-view plane. Because each sensor sees the scene from a different position, fusing their point clouds helps to fill in occluded areas and extend detection range far beyond what a single viewpoint can see. Teams were free to choose any fusion strategy (early, intermediate, or late), and entries were ranked by BEV average precision on a held-out test set.

🔗 Competition & final leaderboard

Congratulations to the team! 🎉