AVOIDDS: Aircraft Vision-based Intruder Detection Dataset and Simulator

TL;DR

AVOIDDS introduces a comprehensive, photorealistic aircraft intruder detection dataset and simulator to advance the development of robust vision-based detect-and-avoid systems for safety-critical aviation applications.

Contribution

This work provides the first large-scale, realistic dataset and integrated simulator for vision-based aircraft intruder detection and collision avoidance, enabling robust ML system development.

Findings

Dataset contains 72,000 photorealistic images under diverse conditions

Provides an evaluation interface for environmental robustness

Includes a closed-loop simulator for downstream collision avoidance

Abstract

Designing robust machine learning systems remains an open problem, and there is a need for benchmark problems that cover both environmental changes and evaluation on a downstream task. In this work, we introduce AVOIDDS, a realistic object detection benchmark for the vision-based aircraft detect-and-avoid problem. We provide a labeled dataset consisting of 72,000 photorealistic images of intruder aircraft with various lighting conditions, weather conditions, relative geometries, and geographic locations. We also provide an interface that evaluates trained models on slices of this dataset to identify changes in performance with respect to changing environmental conditions. Finally, we implement a fully-integrated, closed-loop simulator of the vision-based detect-and-avoid problem to evaluate trained models with respect to the downstream collision avoidance task. This benchmark will enable further research in the design of robust machine learning systems for use in safety-critical applications. The AVOIDDS dataset and code are publicly available at https://purl.stanford.edu/hj293cv5980 and https://github.com/sisl/VisionBasedAircraftDAA respectively.