High-altitude urban object-detection dataset, benchmarks, and deployment evaluation.
Constellation is a 13,000-image benchmark for detecting pedestrians and vehicles in dense urban streetscapes viewed from high-elevation infrastructure cameras.
The dataset captures lighting, weather, shadow, seasonal, and scene changes that are easy to miss in conventional driving datasets. The accompanying evaluations study small-object accuracy, domain-specific pretraining and augmentation, pseudo-labeling, model drift, and inference latency across deployment hardware.
Role: Lead author and dataset lead
Artifacts: Public dataset, model benchmarks, cross-dataset evaluation, and hardware latency tools
We introduce Constellation, a dataset of 13K images suitable for research on detection of objects in dense urban streetscapes observed from high-elevation cameras, collected for a variety of temporal conditions. The dataset addresses the need for curated data to explore problems in small object detection exemplified by the limited pixel footprint of pedestrians observed tens of meters from above. It enables the testing of object detection models for variations in lighting, building shadows, weather, and scene dynamics. We evaluate contemporary object detection architectures on the dataset, observing that state-of-the-art methods have lower performance in detecting small pedestrians compared to vehicles, corresponding to a 10% difference in average precision (AP). Using structurally similar datasets for pretraining the models results in an increase of 1.8% mean AP (mAP). We further find that incorporating domain-specific data augmentations helps improve model performance. Using pseudo-labeled data, obtained from inference outcomes of the best-performing models, improves the performance of the models. Finally, comparing the models trained using the data collected in two different time intervals, we find a performance drift in models due to the changes in intersection conditions over time. The best-performing model achieves a pedestrian AP of 92.0% with 11.5 ms inference time on NVIDIA A100 GPUs, and an mAP of 95.4%.
