Datasets:
-
Add arXiv link and BibTeX citation
-
depth_viz: percentile stretch + NaN-aware filter (test-00004-of-00005.parquet)
-
depth_viz: percentile stretch + NaN-aware filter (test-00003-of-00005.parquet)
-
depth_viz: percentile stretch + NaN-aware filter (test-00002-of-00005.parquet)
-
depth_viz: percentile stretch + NaN-aware filter (test-00001-of-00005.parquet)
-
depth_viz: percentile stretch + NaN-aware filter (test-00000-of-00005.parquet)
-
Trigger viewer reprocess (work around JobManagerCrashedError)
-
Drop dataset_info from YAML (let HF auto-derive from parquet) to fix depth_viz schema mismatch
-
Add depth_viz feature to README
-
Add depth_viz column (test-00004-of-00005.parquet)
-
Add depth_viz column (test-00003-of-00005.parquet)
-
Add depth_viz column (test-00002-of-00005.parquet)
-
Add depth_viz column (test-00001-of-00005.parquet)
-
Add depth_viz column (test-00000-of-00005.parquet)
-
Drop README mentions of raw file tree (folders removed)
-
Drop orphaned raw masks/ folder (data lives in parquet)
-
Drop orphaned raw depth/ folder (data lives in parquet)
-
Drop orphaned raw rgb/ folder (data lives in parquet)
-
Delete orphaned depth_viz/ folder at repo root
-
Drop depth_viz column from README
-
Drop depth_viz column (test-00004-of-00005.parquet)
-
Drop depth_viz column (test-00003-of-00005.parquet)
-
Drop depth_viz column (test-00002-of-00005.parquet)
-
Drop depth_viz column (test-00001-of-00005.parquet)
-
Drop depth_viz column (test-00000-of-00005.parquet)
-
Make dataset title bigger and centered
-
Update README: depth/mask are now 16-bit/8-bit PNG (viewer-friendly)
-
Re-encode depth/mask as PNG (test-00004-of-00005.parquet)
-
Re-encode depth/mask as PNG (test-00003-of-00005.parquet)
-
Re-encode depth/mask as PNG (test-00002-of-00005.parquet)
-
Re-encode depth/mask as PNG (test-00001-of-00005.parquet)
-
Re-encode depth/mask as PNG (test-00000-of-00005.parquet)
-
README: replace Spaces badge with Collection badge (link to prs-eth/pager)
-
Remove placeholder Marigold-era arXiv badge; will be re-added with the PaGeR paper link
-
Switch license from Apache-2.0 to CC BY-4.0 (correct shape for a permissive dataset license; same effective freedoms)
-
Describe parquet schema + add load_dataset example
-
Add Parquet packaging (1 row per panorama, 5 columns)
-
Shorter README, uniform with PanoInfinigen
-
Declare single test split (no train/val)
-
Add depth/
-
Add depth_viz/
-
Add masks/
-
Add rgb/
-
Add dataset card
-
initial commit
