Datasets:
| dataset_info: | |
| features: | |
| - name: image | |
| dtype: image | |
| - name: depth | |
| dtype: image | |
| - name: label | |
| dtype: image | |
| splits: | |
| - name: train | |
| num_bytes: 152339736 | |
| num_examples: 795 | |
| - name: test | |
| num_bytes: 124128749 | |
| num_examples: 654 | |
| download_size: 276518209 | |
| dataset_size: 276468485 | |
| configs: | |
| - config_name: default | |
| data_files: | |
| - split: train | |
| path: data/train-* | |
| - split: test | |
| path: data/test-* | |
| task_categories: | |
| - image-segmentation | |
| # NYUDepth v2 | |
| Easier version for semantic segmentation. | |
| Comes from: | |
| Indoor Segmentation and Support Inference from RGBD Images (ECCV 2012) ([PDF](https://cs.nyu.edu/~fergus/datasets/indoor_seg_support.pdf)) ([Bib](https://cs.nyu.edu/~fergus/datasets/indoor_seg_support.bib)) |