Unreal Engine 5 synthetic-data system for photorealistic urban object detection.
Boundless is a photorealistic synthetic-data pipeline for training object detectors in dense urban streetscapes. It extends the Unreal Engine 5 City Sample into a configurable research system that exports accurately projected 3D annotations across lighting, weather, camera, and scene variations.
The system connects interactive-world building in Unreal Engine with reproducible data-pipeline design for applied computer vision. In cross-domain evaluation, a detector trained on Boundless improved mean average precision by 7.8 points over one trained on CARLA data.
Role: Lead author and system creator
Artifacts: Synthetic datasets for infrastructure and aerial viewpoints, benchmark code, and a reproducible generation methodology
We introduce Boundless, a photo-realistic synthetic data generation system for enabling highly accurate object detection in dense urban streetscapes. Boundless can replace massive real-world data collection and manual ground-truth object annotation (labeling) with an automated and configurable process. Boundless is based on the Unreal Engine 5 (UE5) City Sample project with improvements enabling accurate collection of 3D bounding boxes across different lighting and scene variability conditions. We evaluate the performance of object detection models trained on the dataset generated by Boundless when used for inference on a real-world dataset acquired from medium-altitude cameras. We compare the performance of the Boundless-trained model against the CARLA-trained model and observe an improvement of 7.8 mAP. The results we achieved support the premise that synthetic data generation is a credible methodology for training/fine-tuning scalable object detection models for urban scenes.
