NYU Depth V2: Indoor Scene RGB and Depth Images for Computer Vision

NYU Depth V2 is a dataset of indoor scenes converted from .mat to HDF5 format for deep learning pipelines. The dataset's author, organization, and temporal coverage are not specified in the provided metadata. It is hosted on the Kaggle platform.

Use Cases

• Train depth estimation models based on paired RGB and depth images.
• Develop scene segmentation algorithms based on the provided depth information.
• Benchmark 3D reconstruction pipelines based on the indoor scene data.
• Create synthetic training data for robotics applications based on the depth maps.

Strengths

• Data is specifically formatted for deep learning pipelines, as indicated by the .mat to HDF5 conversion.
• Platform tags suggest the dataset is focused on a core computer vision task: depth estimation.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Row count, file size, and column-level documentation are absent, which may limit suitability assessment.

Provenance

Source

NYU (New York University), inferred from the dataset title.

Collection Method

Likely captured using RGB-D sensors in indoor environments.

Time Range

null

Freshness

Last update date is unknown; freshness unverified.

Geography

null