SkyScenes is a collection of synthetic, densely annotated images designed for aerial scene understanding across a diverse set of environmental conditions. The dataset provides pixel-level semantic labels to address the challenges of data scarcity and annotation costs inherent in real-world aerial photography.
Use Cases
- Train semantic segmentation models using the dense pixel-level labels provided for each synthetic image
- Evaluate model robustness against varying environmental conditions provided in the synthetic scenes
- Develop domain adaptation methods to transfer knowledge from synthetic aerial data to real-world drone datasets
Strengths
- Densely annotated synthetic images for aerial scene understanding
- Features a diverse set of environmental conditions simulated in a controlled framework
- Provides pixel-level semantic labels for fine-grained aerial perception