CM-EVS: A Panoramic RGB-D Dataset for Indoor Scene Understanding

CM-EVS is a curated panoramic RGB-D dataset for indoor scenes, built to maximize geometric coverage with minimal equirectangular frames. The dataset is structured as a redistributable Blender archive and four license-aware adapter packages for local frame regeneration. It was created by anon-cmevs-2026 and last updated on Hugging Face in May 2026.

Use Cases

• Training models for indoor 3D scene reconstruction based on panoramic RGB-D data.
• Developing depth estimation algorithms from equirectangular frames.
• Benchmarking scene understanding tasks like semantic segmentation on panoramic indoor imagery.
• Researching efficient data acquisition strategies for 3D environments based on the coverage-maximization principle.

Strengths

• Dataset is curated under a specific design principle to maximize geometric coverage with few frames.
• The release includes a redistributable Blender archive and adapter packages for license compliance.
• Last updated on Hugging Face on 2026-05-07.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and exact data size are unknown, which may limit suitability assessment.
• The description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Hugging Face user anon-cmevs-2026.

Collection Method

Curated from upstream sources; includes a Blender archive and license-aware adapter packages for local regeneration.

Freshness

Last updated 2026-05-07 01:08:27; freshness should be verified.