Dataset Preview
Duplicate
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed
Error code:   DatasetGenerationError
Exception:    TypeError
Message:      int() argument must be a string, a bytes-like object or a real number, not 'NoneType'
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1539, in _prepare_split_single
                  for key, record in generator:
                                     ^^^^^^^^^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 621, in wrapped
                  for item in generator(*args, **kwargs):
                              ~~~~~~~~~^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/webdataset/webdataset.py", line 131, in _generate_examples
                  for example_idx, example in enumerate(self._get_pipeline_from_tar(tar_path, tar_iterator)):
                                              ~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/webdataset/webdataset.py", line 32, in _get_pipeline_from_tar
                  for filename, f in tar_iterator:
                                     ^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/utils/track.py", line 49, in __iter__
                  for x in self.generator(*self.args):
                           ~~~~~~~~~~~~~~^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/utils/file_utils.py", line 1419, in _iter_from_urlpath
                  with xopen(urlpath, "rb", download_config=download_config, block_size=0) as f:
                       ~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/utils/file_utils.py", line 976, in xopen
                  file_obj = fs.open(paths[0], mode)
                File "<string>", line 3, in open
                File "/usr/local/lib/python3.14/unittest/mock.py", line 1176, in __call__
                  return self._mock_call(*args, **kwargs)
                         ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/unittest/mock.py", line 1180, in _mock_call
                  return self._execute_mock_call(*args, **kwargs)
                         ~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/unittest/mock.py", line 1247, in _execute_mock_call
                  result = effect(*args, **kwargs)
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 794, in wrapped
                  tracker.files[urlpath] = {"read": 0, "size": int(f.size)}
                                                               ~~~^^^^^^^^
              TypeError: int() argument must be a string, a bytes-like object or a real number, not 'NoneType'
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1379, in compute_config_parquet_and_info_response
                  parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
                                                                        ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      builder, max_dataset_size_bytes=max_dataset_size_bytes
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 956, in stream_convert_to_parquet
                  builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
                  ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1401, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                                               ~~~~~~~~~~~~~~~~~~~~~~~~~~^
                      gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  ):
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1579, in _prepare_split_single
                  raise DatasetGenerationError("An error occurred while generating the dataset") from e
              datasets.exceptions.DatasetGenerationError: An error occurred while generating the dataset

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.

b2
unknown
__key__
string
__url__
string
"nqhiMmZyYW1lANIAAABhzwAAAAAAB6mKpBIAVQLTAAAAAAiLgADTAAAAAAAHqHvSAAAAAdIAAAAA0giLgADRADjRADjD2AYAAAA(...TRUNCATED)
carla_difficulty_3000_linear/Town01_Opt/annotation_sim/view_460/view_460_000000_annotation
"hf://datasets/k1ch1/fog-sim-ablation@720612c8b8085aa774db272c6c1dea6d78a5fb6b/carla_difficulty_3000(...TRUNCATED)
"nqhiMmZyYW1lANIAAABhzwAAAAAAB+2YpBIAVQLTAAAAAAiLgADTAAAAAAAH7InSAAAAAdIAAAAA0giLgADRADjRADjD2AYAAAA(...TRUNCATED)
carla_difficulty_3000_linear/Town01_Opt/annotation_sim/view_1228/view_1228_000000_annotation
"hf://datasets/k1ch1/fog-sim-ablation@720612c8b8085aa774db272c6c1dea6d78a5fb6b/carla_difficulty_3000(...TRUNCATED)
"nqhiMmZyYW1lANIAAABhzwAAAAAACCdypBIAVQLTAAAAAAiLgADTAAAAAAAIJmPSAAAAAdIAAAAA0giLgADRADjRADjD2AYAAAA(...TRUNCATED)
carla_difficulty_3000_linear/Town01_Opt/annotation_sim/view_130/view_130_000000_annotation
"hf://datasets/k1ch1/fog-sim-ablation@720612c8b8085aa774db272c6c1dea6d78a5fb6b/carla_difficulty_3000(...TRUNCATED)
"nqhiMmZyYW1lANIAAABhzwAAAAAACJZ3pBIAVQLTAAAAAAiLgADTAAAAAAAIlWjSAAAAAdIAAAAA0giLgADRADjRADjD2AYAAAA(...TRUNCATED)
carla_difficulty_3000_linear/Town01_Opt/annotation_sim/view_2230/view_2230_000000_annotation
"hf://datasets/k1ch1/fog-sim-ablation@720612c8b8085aa774db272c6c1dea6d78a5fb6b/carla_difficulty_3000(...TRUNCATED)
"nqhiMmZyYW1lANIAAABhzwAAAAAACTP3pBIAVQLTAAAAAAiLgADTAAAAAAAJMujSAAAAAdIAAAAA0giLgADRADjRADjD2AYAAAA(...TRUNCATED)
carla_difficulty_3000_linear/Town01_Opt/annotation_sim/view_319/view_319_000000_annotation
"hf://datasets/k1ch1/fog-sim-ablation@720612c8b8085aa774db272c6c1dea6d78a5fb6b/carla_difficulty_3000(...TRUNCATED)
"nqhiMmZyYW1lANIAAABhzwAAAAAACHwCpBIAVQLTAAAAAAiLgADTAAAAAAAIevPSAAAAAdIAAAAA0giLgADRADjRADjD2AYAAAA(...TRUNCATED)
carla_difficulty_3000_linear/Town01_Opt/annotation_sim/view_1335/view_1335_000000_annotation
"hf://datasets/k1ch1/fog-sim-ablation@720612c8b8085aa774db272c6c1dea6d78a5fb6b/carla_difficulty_3000(...TRUNCATED)
"nqhiMmZyYW1lANIAAABhzwAAAAAACOCbpBIAVQLTAAAAAAiLgADTAAAAAAAI34zSAAAAAdIAAAAA0giLgADRADjRADjD2AYAAAA(...TRUNCATED)
carla_difficulty_3000_linear/Town01_Opt/annotation_sim/view_1371/view_1371_000000_annotation
"hf://datasets/k1ch1/fog-sim-ablation@720612c8b8085aa774db272c6c1dea6d78a5fb6b/carla_difficulty_3000(...TRUNCATED)
"nqhiMmZyYW1lANIAAABhzwAAAAAACXCupBIAVQLTAAAAAAiLgADTAAAAAAAJb5/SAAAAAdIAAAAA0giLgADRADjRADjD2AYAAAA(...TRUNCATED)
carla_difficulty_3000_linear/Town01_Opt/annotation_sim/view_2059/view_2059_000000_annotation
"hf://datasets/k1ch1/fog-sim-ablation@720612c8b8085aa774db272c6c1dea6d78a5fb6b/carla_difficulty_3000(...TRUNCATED)
"nqhiMmZyYW1lANIAAABhzwAAAAAACJ9TpBIAVQLTAAAAAAiLgADTAAAAAAAInkTSAAAAAdIAAAAA0giLgADRADjRADjD2AYAAAA(...TRUNCATED)
carla_difficulty_3000_linear/Town01_Opt/annotation_sim/view_2422/view_2422_000000_annotation
"hf://datasets/k1ch1/fog-sim-ablation@720612c8b8085aa774db272c6c1dea6d78a5fb6b/carla_difficulty_3000(...TRUNCATED)
"nqhiMmZyYW1lANIAAABhzwAAAAAACFZrpBIAVQLTAAAAAAiLgADTAAAAAAAIVVzSAAAAAdIAAAAA0giLgADRADjRADjD2AYAAAA(...TRUNCATED)
carla_difficulty_3000_linear/Town01_Opt/annotation_sim/view_332/view_332_000000_annotation
"hf://datasets/k1ch1/fog-sim-ablation@720612c8b8085aa774db272c6c1dea6d78a5fb6b/carla_difficulty_3000(...TRUNCATED)
End of preview.

Fog-Sim Ablation

Two ablations of a simulated full-waveform LiDAR (FWL) dataset of CARLA scenes with a smoke plume between the sensor and cars/pedestrians. Each ablation removes one ingredient of the simulator, so that its effect on fog/object segmentation (and on transfer to real data) can be measured:

variant smoke density receiver model
carla_difficulty_3000_uniform uniform inside the plume (same shape, extent and wind as the baseline; only the inner density variation is removed) SPAD receiver (latched cells, dead time 3 bins, per-row ceiling, random photon draw)
carla_difficulty_3000_linear ("no-SPAD") same as the baseline none: each bin holds the expected number of photons after the laser pulse; no latch, no ceiling, no random draw

The baseline dataset (carla_difficulty_3000, inhomogeneous plume + SPAD receiver) is not part of this repository. Both ablations reuse the baseline's scenes, sensor poses, actors and plume parameters view by view (ablation_of in each view's meta).

Contents

  • 8 CARLA towns: Town01, 02, 03, 04, 05, 06, 07, 10HD (*_Opt).

  • 1,500 views per town, one frame each → 12,000 frames per variant.

  • 3 difficulty levels, 500 views each, set by the fraction of object signal that survives the smoke. The smoke extinction σt is solved per view to hit the target.

    difficulty view ids surviving-fraction target
    easy view_000–view_499 0.80
    normal view_1000–view_1499 0.50
    hard view_2000–view_2499 0.25
  • In the uniform variant, σt has to be 1.3–1.6× the baseline's to reach the same surviving fraction.

Recommended split

Train on Town01/02/03/06/07/10HD, validate on Town04 and test on Town05.

Layout

Every folder is packed as one tar per town (all files < 50 GB). Run tar xf at the same directory level to restore the original paths, e.g.

carla_difficulty_3000_uniform/
  Town01_Opt/
    dataset_map.json                 # list of views with actors, σt and relative paths
    meta.tar                         # meta/view_XXX.json, meta/view_XXX.frames/000000.json
    annotation_sim.tar               # labels from the simulator
    gt_annotation_sim.tar            # object labels of the smoke-free render
    annotation_v6_sim.tar            # (uniform only) v6 labels, one bin per peak
    annotation_v6_sim_expand.tar     # (uniform only) v6 labels expanded over each peak
carla_difficulty_3000_uniform_gain/
  Town01_Opt/
    data.tar                         # gain-normalised waveforms
carla_difficulty_3000_linear/        # same as above, without the v6 folders
carla_difficulty_3000_linear_gain/
huggingface-cli download k1ch1/fog-sim-ablation --repo-type dataset \
  --include "carla_difficulty_3000_uniform*/Town05_Opt/*" --local-dir .
# tar members carry the full path (carla_difficulty_3000_*/TownXX_Opt/...),
# so extract from the top directory:
find . -name '*.tar' -exec tar xf {} \;

Not included: the smoke-free (clear) waveforms, raw photon counts, and the cuboids/ folder referenced in dataset_map.json.

File format

All arrays are blosc2-packed NumPy arrays (.b2) with shape (400, 512, 700) = (azimuth, elevation, time bin).

  • Azimuth step 0.1°, elevation step 0.05°, time bin 1 ns (≈ 0.15 m). Range zero is at bin 9.
  • Waveform path: data/<view>/<view>_000000.b2.
  • Label path: <annotation folder>/<view>/<view>_000000_annotation.b2.
import blosc2

wave = blosc2.unpack_array2(open("view_000_000000.b2", "rb").read())             # float16, 0–1
label = blosc2.unpack_array2(open("view_000_000000_annotation.b2", "rb").read())  # int8
ray = wave[x, y, :]   # one waveform

Waveforms (*_gain/.../data): float16 in [0, 1], gain-normalised and clipped at 1.

  • uniform: SPAD counts divided by the per-row trial count (18 × shots).
  • linear: expected photons passed through the same gain step and clipped at 1. They are not derived from integer counts.

Labels: int8, 0 = none/unknown, 1 = object, 4 = fog.

folder content
annotation_sim labels produced by the simulator from the rendered transient. In the linear variant they are computed on the expected photons, so their definition differs from the baseline.
gt_annotation_sim object labels of the same view rendered without smoke (0/1 only).
annotation_v6_sim (uniform) peak-based v6 labels: one bin per detected peak.
annotation_v6_sim_expand (uniform) v6 labels expanded over the whole peak. Used for training.

For the linear variant, training used the baseline's annotation_v6_sim_expand. Scenes, actors and smoke are identical to the baseline view by view, so those labels apply. They are not included here.

Meta (meta/<view>.json):

  • sensor pose and grid
  • actors (category, asset, pose, size, visible rays, whether behind smoke)
  • smoke-plume parameters, including uniform and the solved sigma_t
  • difficulty target and the predicted surviving fraction
  • receiver model and seeds

Per-frame meta (meta/<view>.frames/000000.json) holds the smoke source, the smoke bounds and the fog label bin range.

Known stale strings in the metadata:

  • The uniform variant's sensor_model.pulse reads "gamma t^3, 7 bins FWHM". The pulse actually used is t² with a 5-bin FWHM, the same as the linear variant.
  • dataset_map.json → layout.data still describes the files as "uint8 photon counts". The released *_gain files are float16 gain-normalised values.
  • dataset_root in dataset_map.json is an absolute path on the authors' machine. Use the relative paths in views[*] instead.

Generation

Generated with a Mitsuba 3 transient renderer over CARLA (UE) scenes (FWL_Mitsuba, generate_ue_dataset.py, seed 20260930):

  • Smoke is a procedural plume.
  • The pulse shape is t²·exp(−t/a) with a 5-bin FWHM.
  • Photon scale is 20, with no background light.
Downloads last month
60