ClaraVid: Aerial Scene Reconstruction Benchmark with 16,917 Frames

ClaraVid is a synthetic dataset for semantic and geometric neural reconstruction from low altitude UAV imagery. It contains 16,917 multimodal frames collected across 8 UAV missions over diverse environments. The dataset was created by radubeche and was last updated on October 31, 2025.

Use Cases

• Training neural networks for 3D scene reconstruction based on low-altitude aerial imagery.
• Benchmarking semantic segmentation algorithms for urban and natural environments from a UAV perspective.
• Developing multimodal fusion models based on the described multimodal frame data.
• Researching complexity profiling for aerial scenes based on the delentropy-based method mentioned.

Strengths

• Contains 16,917 multimodal frames, providing a substantial volume of data.
• Covers 8 distinct UAV missions, suggesting diversity in flight paths and conditions.
• Includes diverse environments such as urban and urban high-rise areas.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data is synthetic, which may limit its applicability to real-world deployment scenarios.

Provenance

Source

huggingface

Collection Method

Synthetically generated for UAV/aerial imagery reconstruction tasks.

Freshness

Last updated 2025-10-31 04:51:31; freshness should be verified.