Boundless: Generating Photorealistic Synthetic Data for Object Detection in Urban Streetscapes
This addresses the need for scalable and automated data generation for object detection in urban scenes, offering a credible alternative to costly real-world data collection.
The paper tackles the problem of training object detection models for urban streetscapes by introducing Boundless, a photorealistic synthetic data generation system that replaces real-world data collection and manual annotation, resulting in a 7.8 mAP improvement over a CARLA-trained model on real-world data.
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.