Datasets:

milleleaf
/

hk-building-textures

Modalities:

Geospatial

Tags:

License:

Dataset card Files Files and versions

xet

Community

Dataset Viewer

Duplicate

Split (1)

train

The dataset viewer is not available for this split.

Cannot load the dataset split (in streaming mode) to extract the first rows.

Error code:   StreamingRowsError
Exception:    TypeError
Message:      Couldn't cast array of type
struct<1-SE-19D: struct<product: string, sheet: string, rev: int64, code: string, fetch: struct<threads: int64, span_MB: double, pause_s: double, stop_on_403: bool, requests: int64, http403: int64, first_403_utc: null, MB_before_403: null, MB: double, error: null>, zipcut: struct<zip_MB: double, members: int64, fetched_members: int64, MB: double, downloaded_MB: double, requests: int64>, buildings: int64, seconds: struct<fetch: int64, upload: int64>, peak_rss_GB: double, files: int64, bytes: int64>, 1-SE-20C: struct<product: string, sheet: string, rev: int64, code: string, fetch: struct<threads: int64, span_MB: double, pause_s: double, stop_on_403: bool, requests: int64, http403: int64, first_403_utc: null, MB_before_403: null, MB: double, error: null>, zipcut: struct<zip_MB: double, members: int64, fetched_members: int64, MB: double, downloaded_MB: double, requests: int64>, convert: struct<buildings: int64, triangles: int64, images: int64, imageBytesIn: int64, imageBytesOut: int64, glbBytes: int64, cells: int64, tex: string, ge: int64, MB: int64, imageMB_in: double, imageMB_out: int64>, buildings: int64, seconds: struct<fetch: int64, convert: int64, upload: int64>, peak_rss_GB: double, files: int64, bytes: int64, region: list<item: double>>, 1-SE-24A: struct<product: string, sheet: string, rev: int64, code: string, fetch: struct<threads: int64, span_MB: double, pause_s: double, stop_on_403: bool, requests: int64, http403: int64, first_403_utc: null, MB_before_403: null, MB: d
...
<product: string, sheet: string, rev: int64, code: string, fetch: struct<threads: int64, span_MB: double, pause_s: double, stop_on_403: bool, requests: int64, http403: int64, first_403_utc: null, MB_before_403: null, MB: double, error: null>, zipcut: struct<zip_MB: double, members: int64, fetched_members: int64, MB: double, downloaded_MB: double, requests: int64>, convert: struct<buildings: int64, triangles: int64, images: int64, imageBytesIn: int64, imageBytesOut: int64, glbBytes: int64, cells: int64, tex: string, ge: int64, MB: double, imageMB_in: double, imageMB_out: int64>, buildings: int64, seconds: struct<fetch: int64, convert: int64, upload: int64>, peak_rss_GB: double, files: int64, bytes: int64, region: list<item: double>>, 8-SW-8A: struct<product: string, sheet: string, rev: int64, code: string, fetch: struct<threads: int64, span_MB: double, pause_s: double, stop_on_403: bool, requests: int64, http403: int64, first_403_utc: null, MB_before_403: null, MB: double, error: null>, zipcut: struct<zip_MB: double, members: int64, fetched_members: int64, MB: double, downloaded_MB: double, requests: int64>, convert: struct<buildings: int64, triangles: int64, images: int64, imageBytesIn: int64, imageBytesOut: int64, glbBytes: int64, cells: int64, tex: string, ge: int64, MB: double, imageMB_in: int64, imageMB_out: double>, buildings: int64, seconds: struct<fetch: int64, convert: int64, upload: int64>, peak_rss_GB: double, files: int64, bytes: int64, region: list<item: double>>>
to
{'11-NW-14C': {'product': Value('string'), 'sheet': Value('string'), 'rev': Value('int64'), 'code': Value('string'), 'zipcut': {'11-NW-13B': {'zip_MB': Value('float64'), 'members': Value('int64'), 'fetched_members': Value('int64'), 'MB': Value('float64'), 'downloaded_MB': Value('float64'), 'requests': Value('int64'), 'tilesets': Value('int64'), 'leaf_tiles': Value('int64'), 'missing_in_zip': Value('int64')}, '11-NW-13D': {'zip_MB': Value('float64'), 'members': Value('int64'), 'fetched_members': Value('int64'), 'MB': Value('float64'), 'downloaded_MB': Value('float64'), 'requests': Value('int64'), 'tilesets': Value('int64'), 'leaf_tiles': Value('int64'), 'missing_in_zip': Value('int64')}, '11-NW-14A': {'zip_MB': Value('float64'), 'members': Value('int64'), 'fetched_members': Value('int64'), 'MB': Value('float64'), 'downloaded_MB': Value('float64'), 'requests': Value('int64'), 'tilesets': Value('int64'), 'leaf_tiles': Value('int64'), 'missing_in_zip': Value('int64')}, '11-NW-14B': {'zip_MB': Value('float64'), 'members': Value('int64'), 'fetched_members': Value('int64'), 'MB': Value('float64'), 'downloaded_MB': Value('float64'), 'requests': Value('int64'), 'tilesets': Value('int64'), 'leaf_tiles': Value('int64'), 'missing_in_zip': Value('int64')}, '11-NW-14C': {'zip_MB': Value('float64'), 'members': Value('int64'), 'fetched_members': Value('int64'), 'MB': Value('float64'), 'downloaded_MB': Value('float64'), 'requests': Value('int64'), 'tilesets': Value('int64'), 'leaf_tiles': Val
...
hts': Value('int64'), 'dz': Value('int64'), 'chart_deg': Value('int64'), 'uncovered': List({'id': Value('string'), 'type': Value('string'), 'top': Value('float64'), 'cells': Value('int64'), 'cov': Value('int64'), 'any': Value('int64')}), 'baked': Value('int64'), 'outside_mesh': Value('int64'), 'MB': Value('float64'), 'MB_geometry_raw': Value('float64'), 'MB_geometry_meshopt': Value('float64'), 'MB_texture': Value('float64'), 'tris': Value('int64'), 'tris_degenerate_dropped': Value('int64'), 'twin_tris_dropped': Value('int64'), 'charts': Value('int64'), 'charts_twin_dropped': Value('int64'), 'charts_tiny_lt_0_25m2': Value('int64'), 'charts_hidden': Value('int64'), 'atlas_fill': Value('float64'), 'kB_per_model_median': Value('int64'), 'ms_per_model_median': Value('int64'), 'coarser_texel': Value('int64'), 'texels_M': Value('float64'), 'hidden_texels_M': Value('float64'), 'wall_nohit': Value('float64'), 'party_wall_share': Value('float64'), 'party_nohit': Value('float64'), 'street_wall_nohit': Value('float64'), 'roof_nohit': Value('float64'), 'down_nohit': Value('float64'), 'offset_wall_m_p10_p50_p90': List(Value('float64')), 'offset_roof_m_p10_p50_p90': List(Value('float64'))}}, 'buildings': Value('int64'), 'seconds': {'footprints': Value('int64'), 'fetch_photomesh': Value('int64'), 'fetch_3dsd': Value('int64'), 'bake': Value('int64'), 'upload': Value('int64')}, 'peak_rss_GB': Value('float64'), 'files': Value('int64'), 'bytes': Value('int64'), 'region': List(Value('float64'))}}
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 147, in get_rows_or_raise
                  return get_rows(
                      dataset=dataset,
                  ...<4 lines>...
                      column_names=column_names,
                  )
                File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
                  return func(*args, **kwargs)
                File "/src/services/worker/src/worker/utils.py", line 127, in get_rows
                  rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
                File "/src/services/worker/src/worker/utils.py", line 483, in safe_iter
                  yield from ds.decode(False) if ds.features else ds
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2951, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2461, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2486, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 547, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 430, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2378, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2312, in cast_table_to_schema
                  cast_array_to_feature(
                  ~~~~~~~~~~~~~~~~~~~~~^
                      table[name] if name in table_column_names else pa.array([None] * len(table), type=schema.field(name).type),
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                      feature,
                      ^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 1861, in wrapper
                  return pa.chunked_array([func(chunk, *args, **kwargs) for chunk in array.chunks])
                                           ~~~~^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2158, in cast_array_to_feature
                  raise TypeError(f"Couldn't cast array of type\n{_short_str(array.type)}\nto\n{_short_str(feature)}")
              TypeError: Couldn't cast array of type
              struct<1-SE-19D: struct<product: string, sheet: string, rev: int64, code: string, fetch: struct<threads: int64, span_MB: double, pause_s: double, stop_on_403: bool, requests: int64, http403: int64, first_403_utc: null, MB_before_403: null, MB: double, error: null>, zipcut: struct<zip_MB: double, members: int64, fetched_members: int64, MB: double, downloaded_MB: double, requests: int64>, buildings: int64, seconds: struct<fetch: int64, upload: int64>, peak_rss_GB: double, files: int64, bytes: int64>, 1-SE-20C: struct<product: string, sheet: string, rev: int64, code: string, fetch: struct<threads: int64, span_MB: double, pause_s: double, stop_on_403: bool, requests: int64, http403: int64, first_403_utc: null, MB_before_403: null, MB: double, error: null>, zipcut: struct<zip_MB: double, members: int64, fetched_members: int64, MB: double, downloaded_MB: double, requests: int64>, convert: struct<buildings: int64, triangles: int64, images: int64, imageBytesIn: int64, imageBytesOut: int64, glbBytes: int64, cells: int64, tex: string, ge: int64, MB: int64, imageMB_in: double, imageMB_out: int64>, buildings: int64, seconds: struct<fetch: int64, convert: int64, upload: int64>, peak_rss_GB: double, files: int64, bytes: int64, region: list<item: double>>, 1-SE-24A: struct<product: string, sheet: string, rev: int64, code: string, fetch: struct<threads: int64, span_MB: double, pause_s: double, stop_on_403: bool, requests: int64, http403: int64, first_403_utc: null, MB_before_403: null, MB: d
              ...
              <product: string, sheet: string, rev: int64, code: string, fetch: struct<threads: int64, span_MB: double, pause_s: double, stop_on_403: bool, requests: int64, http403: int64, first_403_utc: null, MB_before_403: null, MB: double, error: null>, zipcut: struct<zip_MB: double, members: int64, fetched_members: int64, MB: double, downloaded_MB: double, requests: int64>, convert: struct<buildings: int64, triangles: int64, images: int64, imageBytesIn: int64, imageBytesOut: int64, glbBytes: int64, cells: int64, tex: string, ge: int64, MB: double, imageMB_in: double, imageMB_out: int64>, buildings: int64, seconds: struct<fetch: int64, convert: int64, upload: int64>, peak_rss_GB: double, files: int64, bytes: int64, region: list<item: double>>, 8-SW-8A: struct<product: string, sheet: string, rev: int64, code: string, fetch: struct<threads: int64, span_MB: double, pause_s: double, stop_on_403: bool, requests: int64, http403: int64, first_403_utc: null, MB_before_403: null, MB: double, error: null>, zipcut: struct<zip_MB: double, members: int64, fetched_members: int64, MB: double, downloaded_MB: double, requests: int64>, convert: struct<buildings: int64, triangles: int64, images: int64, imageBytesIn: int64, imageBytesOut: int64, glbBytes: int64, cells: int64, tex: string, ge: int64, MB: double, imageMB_in: int64, imageMB_out: double>, buildings: int64, seconds: struct<fetch: int64, convert: int64, upload: int64>, peak_rss_GB: double, files: int64, bytes: int64, region: list<item: double>>>
              to
              {'11-NW-14C': {'product': Value('string'), 'sheet': Value('string'), 'rev': Value('int64'), 'code': Value('string'), 'zipcut': {'11-NW-13B': {'zip_MB': Value('float64'), 'members': Value('int64'), 'fetched_members': Value('int64'), 'MB': Value('float64'), 'downloaded_MB': Value('float64'), 'requests': Value('int64'), 'tilesets': Value('int64'), 'leaf_tiles': Value('int64'), 'missing_in_zip': Value('int64')}, '11-NW-13D': {'zip_MB': Value('float64'), 'members': Value('int64'), 'fetched_members': Value('int64'), 'MB': Value('float64'), 'downloaded_MB': Value('float64'), 'requests': Value('int64'), 'tilesets': Value('int64'), 'leaf_tiles': Value('int64'), 'missing_in_zip': Value('int64')}, '11-NW-14A': {'zip_MB': Value('float64'), 'members': Value('int64'), 'fetched_members': Value('int64'), 'MB': Value('float64'), 'downloaded_MB': Value('float64'), 'requests': Value('int64'), 'tilesets': Value('int64'), 'leaf_tiles': Value('int64'), 'missing_in_zip': Value('int64')}, '11-NW-14B': {'zip_MB': Value('float64'), 'members': Value('int64'), 'fetched_members': Value('int64'), 'MB': Value('float64'), 'downloaded_MB': Value('float64'), 'requests': Value('int64'), 'tilesets': Value('int64'), 'leaf_tiles': Value('int64'), 'missing_in_zip': Value('int64')}, '11-NW-14C': {'zip_MB': Value('float64'), 'members': Value('int64'), 'fetched_members': Value('int64'), 'MB': Value('float64'), 'downloaded_MB': Value('float64'), 'requests': Value('int64'), 'tilesets': Value('int64'), 'leaf_tiles': Val
              ...
              hts': Value('int64'), 'dz': Value('int64'), 'chart_deg': Value('int64'), 'uncovered': List({'id': Value('string'), 'type': Value('string'), 'top': Value('float64'), 'cells': Value('int64'), 'cov': Value('int64'), 'any': Value('int64')}), 'baked': Value('int64'), 'outside_mesh': Value('int64'), 'MB': Value('float64'), 'MB_geometry_raw': Value('float64'), 'MB_geometry_meshopt': Value('float64'), 'MB_texture': Value('float64'), 'tris': Value('int64'), 'tris_degenerate_dropped': Value('int64'), 'twin_tris_dropped': Value('int64'), 'charts': Value('int64'), 'charts_twin_dropped': Value('int64'), 'charts_tiny_lt_0_25m2': Value('int64'), 'charts_hidden': Value('int64'), 'atlas_fill': Value('float64'), 'kB_per_model_median': Value('int64'), 'ms_per_model_median': Value('int64'), 'coarser_texel': Value('int64'), 'texels_M': Value('float64'), 'hidden_texels_M': Value('float64'), 'wall_nohit': Value('float64'), 'party_wall_share': Value('float64'), 'party_nohit': Value('float64'), 'street_wall_nohit': Value('float64'), 'roof_nohit': Value('float64'), 'down_nohit': Value('float64'), 'offset_wall_m_p10_p50_p90': List(Value('float64')), 'offset_roof_m_p10_p50_p90': List(Value('float64'))}}, 'buildings': Value('int64'), 'seconds': {'footprints': Value('int64'), 'fetch_photomesh': Value('int64'), 'fetch_3dsd': Value('int64'), 'bake': Value('int64'), 'upload': Value('int64')}, 'peak_rss_GB': Value('float64'), 'files': Value('int64'), 'bytes': Value('int64'), 'region': List(Value('float64'))}}

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

Hong Kong buildings with photo textures (3D Tiles)

Hong Kong buildings with a real photo texture, in two forms, built per 1:1000 map sheet by milleleaf/hk-building-textures from Lands Department open data. The tiles are in the bucket milleleaf/hk-building-textures and are read in place by browsers (CORS and HTTP Range). This dataset repo holds this card and the two manifests.

A work in progress, sheet by sheet: on 2026-10-23, 12 of the 3,456 sheets were published as indiv-k4 (6,152 buildings) and 8 of 3,455 as bake3dsd (4,491 buildings). Each product's manifest.json has the current count (sheets_published of sheets_in_index).

Product	For	What it is
`indiv-k4/`	desktop	Lands D's own Individualised models (one glTF per building, photo textures), textures re-encoded as KTX2 ETC1S at 1/4 size, geometry meshopt-compressed. Buildings are grouped in 150 m cells that load within about 2 km of the camera.
`bake3dsd/`	phones and tablets	Lands D's 3D Spatial Data building models (simple shapes), each given its own texture atlas baked from the Open3Dhk photomesh at 20 cm texels (KTX2 ETC1S, meshopt). About ten times smaller than `indiv-k4`.

Layout, per product: tileset.json (root, 3D Tiles 1.1: 1:5000 map-sheet groups, each map sheet an external tileset), <sheet>/tileset.json with one .glb per building, <sheet>/report.json (counts, sizes, timings of that sheet's build), manifest.json (every sheet's report and the totals). Both need a KTX2 (Basis Universal) and a meshopt decoder, e.g. three.js KTX2Loader and MeshoptDecoder with 3d-tiles-renderer, or CesiumJS.

const tiles = new TilesRenderer('https://huggingface.co/buckets/milleleaf/hk-building-textures/resolve/bake3dsd/tileset.json')

Known defects

bake3dsd (our bake; seen on the Mong Kok pilot sheets, baker of hk-ifcx commit 6b7e5bf, which reduced but did not remove them):

trees, signs and canopies in front of a wall can still be baked onto the wall;
faces the photomesh does not see (narrow lanes, under podiums) take another part of the model's texture or its mean colour, so they look flat;
on the pilot sheets 9 to 23 % of wall texels and 11 to 24 % of roof texels found no photomesh surface and were filled (per-sheet figures in each report.json); curved roofs come out pale;
a CSDI footprint with no 3D Spatial Data model and no heights is not drawn.

Both products: heights are metres above Hong Kong Principal Datum, used as ellipsoidal heights without conversion, and indiv-k4 ignores the HK1980 grid convergence (about 0.003 degrees). Anonymous Hugging Face downloads are limited to 3,000 files per 5 minutes per IP address. One street-level view of Mong Kok loads about 1,400 files of indiv-k4 or 2,350 of bake3dsd, so a second view within 5 minutes fails part-way (HTTP 429) unless the files are served through a cache.

Sources and terms

© Lands Department, HKSAR Government. Data from data.map.gov.hk and the CSDI Portal (portal.csdi.gov.hk), free for commercial and non-commercial use with attribution, under the CSDI terms. The Lands Department owns the intellectual property rights in the data. This dataset is derived from it, provided as is with no warranty, and not endorsed by the Government. Full source list: the GitHub repository's sources.json.

Downloads last month: -

Total file size:

284 kB