qizhangslam/scannet-bags146-paired-recovery-20261011
ScanNet BAGS paired recovery Incremental recovery of the 146-scene, 3,102-frame selection in zhaoshiwen/newscannet. This repository adds source RGB, sensor depth, native ScanNet GT poses, calibration, mesh, and newly measured NIMA-KONIQ and LPIPS to the surviving BAGS renders. Recovered selection complete: 146 scenes and 3,102 frames. Consult coverage.json for the scenes and frames actually published. The complete target whitelist is provenance/target_frame_ids.json. The old… See the full description on the dataset page: https://huggingface.co/datasets/qizhangslam/scannet-bags146-paired-recove
| Author / organization | qizhangslam |
|---|---|
| License (as declared) | other |
| Downloads | 1.5K (as reported by Hugging Face) |
| Likes | 0 |
| Created | 11 October 2026 |
| Last updated at source | 11 October 2026 |
Popularity, last 2 days
1,472 now (+5 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
More from qizhangslam
- qizhangslam/blur-slam-exp-best-preservation-s1Dataset
- qizhangslam/unblur-tum-full-results-20260928Dataset
- qizhangslam/e2e-stream-slam-training-datasetDataset
- qizhangslam/six-method-five-category-aligned-20260915Dataset
- qizhangslam/scannet-bags146-recovery-archive-20261011Dataset
- qizhangslam/unblur-slam-two-stage-evidence-20260910-publicDataset
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.