DeSOPE: RGB Image Sequences with Object Pose and 3D Models

Visioner's DeSOPE dataset provides RGB image sequences paired with object pose annotations, camera intrinsics, and 3D mesh models. The data is organized into scenes containing static sequences with pose files and corresponding image frames. The dataset was last updated on June 8, 2026.

Use Cases

• Training 6D object pose estimation models based on the provided pose annotations and RGB images.
• Developing 3D object reconstruction pipelines using the supplied canonical and deformed mesh models.
• Benchmarking camera calibration and scene geometry algorithms based on the camera intrinsics data.
• Creating synthetic data or augmenting real data for vision tasks using the provided 3D assets.

Strengths

• Includes multiple data modalities: RGB images, object pose annotations, camera intrinsics, and 3D mesh models.
• Provides both canonical and deformed versions of object models, which may support research into object deformation.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

visioner

Freshness

Last updated 2026-06-08 08:45:05; freshness should be verified.