OccuFly: Aerial Imagery for Semantic Scene Completion and Depth Estimation

OccuFly is a real-world, large-scale camera-based benchmark for Semantic Scene Completion and Metric Monocular Depth Estimation from an aerial perspective. The dataset was released by author markus-42 following its acceptance as a CVPR 2026 Oral presentation. Associated resources include a GitHub repository, project page, and a paper available on arXiv.

Use Cases

• Training models for semantic scene completion based on aerial camera images.
• Benchmarking monocular depth estimation algorithms from an aerial perspective.
• Developing and evaluating 3D reconstruction techniques for outdoor environments.

Strengths

• Described as the first real-world, large-scale camera-based benchmark for its specific tasks.
• Accepted as a CVPR 2026 Oral, suggesting peer-reviewed quality.
• Associated with a published paper (arXiv:2512.20770) and detailed project documentation.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and dataset size are unknown, which may limit suitability assessment.
• Data may reflect geographic or temporal bias inherent to its unspecified collection sources.

Provenance

Source

markus-42

Collection Method

Likely collected from real-world aerial camera sources, though the specific method is not detailed.

Freshness

Last updated 2026-06-01 16:27:08; freshness should be verified.