Scalable Behavior Cloning with Open Data, Training, and Evaluation
Paper • 2606.27375 • Published • 1
Dataset Viewer
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: ValueError
Message: Expected object or value
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 290, in _generate_tables
pa_table = paj.read_json(
io.BytesIO(batch), read_options=paj.ReadOptions(block_size=block_size)
)
File "pyarrow/_json.pyx", line 342, in pyarrow._json.read_json
File "pyarrow/error.pxi", line 155, in pyarrow.lib.pyarrow_internal_check_status
File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
raise convert_status(status)
pyarrow.lib.ArrowInvalid: JSON parse error: Column() changed from object to number in row 0
During handling of the above exception, another exception occurred:
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 2840, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2373, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2398, 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 536, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 419, 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 304, in _generate_tables
batch = json_encode_fields_in_json_lines(original_batch, json_field_paths)
File "/usr/local/lib/python3.14/site-packages/datasets/utils/json.py", line 111, in json_encode_fields_in_json_lines
examples = [ujson_loads(line) for line in original_batch.splitlines()]
~~~~~~~~~~~^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/utils/json.py", line 20, in ujson_loads
return pd.io.json.ujson_loads(*args, **kwargs)
~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
ValueError: Expected object or valueNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Classical optical flow on selected ABC-130k intervals
Precomputed OpenCV DIS and Farneback backward optical-flow estimates for five eight-second intervals from XDOF/ABC-130k, using three 640x480 camera streams. These reusable motion fields are estimates, not ground-truth correspondence. This release contains 135 arrays and no learned model weights.
Files and interpretation
flow/T01/{top,left,right}/...npy (and T02 through T05) contains float32 arrays
of shape (240, 480, 640, 2). The final axis is (dx, dy) in full-resolution
pixels. At current pixel (x,y), the corresponding previous-image coordinate
is (x+dx,y+dy). The first frame is zero. Each array is approximately 590 MB.
oracle_dis.npy: DIS estimated directly from full-resolution source images; an original-input diagnostic, not ground truth.s2_...ands4_...: point subsampling by 2 or 4 per axis followed by phase-aligned bicubic upsampling before estimating flow at 640x480.fixed: sampling phase(0,0);jittered: cycle all s*s phases, x fastest.dis: OpenCV DIS medium preset, spatial propagation enabled.farneback: pyr_scale=.5, levels=4, winsize=21, iterations=4, poly_n=7, poly_sigma=1.5, flags=0. Flow inputs are RGB converted to grayscale.
Original camera frames were held onto a 30 Hz time grid. Common grid ticks are
not simultaneous camera exposures; repeated exposures must be considered when
using the arrays. source_provenance.json records original episodes, revision,
frame mappings and timestamps. manifest.json records byte sizes and SHA-256.
Download one array
from huggingface_hub import hf_hub_download
import numpy as np
path = hf_hub_download(
repo_id="HuggingFacer112358/abc130k-classical-optical-flow-20260922",
repo_type="dataset", filename="flow/T01/top/s2_fixed_dis.npy",
revision="<pin the desired commit hash>",
)
flow = np.load(path, mmap_mode="r", allow_pickle=False)
Attribution and license
Derived from XDOF/ABC-130k,
revision 75ca0b88bda489f2bd935d72454593ecb90efb52, whose dataset card declares
Apache-2.0. These files are transformed optical-flow derivatives produced by
Dreamscale Labs, not the original dataset or XDOF-provided flow annotations.
The Apache-2.0 license is included as LICENSE. Cite the original dataset:
Allshire et al., Scalable Behavior Cloning with Open Data, Training, and
Evaluation, 2026, arXiv:2606.27375. This derived release does not imply XDOF
endorsement. No raw source video is redistributed here.
- Downloads last month
- 240