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:    ArrowInvalid
Message:      Failed to parse string: 'no' as a scalar of type double
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1837, in _prepare_split_single
                  writer.write_table(table)
                  ~~~~~~~~~~~~~~~~~~^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 765, in write_table
                  self._write_table(pa_table, writer_batch_size=writer_batch_size)
                  ~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 773, in _write_table
                  pa_table = table_cast(pa_table, self._schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2369, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2303, 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 1852, 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 2143, in cast_array_to_feature
                  return array_cast(
                      array,
                  ...<2 lines>...
                      allow_decimal_to_str=allow_decimal_to_str,
                  )
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 1854, in wrapper
                  return func(array, *args, **kwargs)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2006, in array_cast
                  return array.cast(pa_type)
                         ~~~~~~~~~~^^^^^^^^^
                File "pyarrow/array.pxi", line 1147, in pyarrow.lib.Array.cast
                File "/usr/local/lib/python3.14/site-packages/pyarrow/compute.py", line 412, in cast
                  return call_function("cast", [arr], options, memory_pool)
                File "pyarrow/_compute.pyx", line 604, in pyarrow._compute.call_function
                File "pyarrow/_compute.pyx", line 399, in pyarrow._compute.Function.call
                  result = GetResultValue(
                File "pyarrow/error.pxi", line 155, in pyarrow.lib.pyarrow_internal_check_status
                  return check_status(status)
                File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
                  raise convert_status(status)
              pyarrow.lib.ArrowInvalid: Failed to parse string: 'no' as a scalar of type double
              
              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 1369, 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 948, 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 1683, 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 1869, 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.

frame_id
int64
bbox_left
float64
bbox_top
float64
bbox_width
float64
bbox_height
float64
has_bbox
bool
source_marked_invalid
bool
invalid_note
string
occlusion_raw
float64
similar_raw
string
1
411
404
24
11
true
false
null
1
no
2
413
404
25
13
true
false
null
1
no
3
417
405
22
12
true
true
3-65
1
no
4
422
407
19
11
true
true
null
1
no
5
422
408
20
10
true
true
null
1
no
6
426
408
17
11
true
true
null
1
no
7
428
409
12
9
true
true
null
1
no
8
431
409
12
8
true
true
null
1
no
9
433
410
8
7
true
true
null
1
no
10
null
null
null
null
false
true
null
null
null
11
null
null
null
null
false
true
null
null
null
12
null
null
null
null
false
true
null
null
null
13
null
null
null
null
false
true
null
null
null
14
null
null
null
null
false
true
null
null
null
15
null
null
null
null
false
true
null
null
null
16
null
null
null
null
false
true
null
null
null
17
null
null
null
null
false
true
null
null
null
18
null
null
null
null
false
true
null
null
null
19
null
null
null
null
false
true
null
null
null
20
null
null
null
null
false
true
null
null
null
21
null
null
null
null
false
true
null
null
null
22
null
null
null
null
false
true
null
null
null
23
null
null
null
null
false
true
null
null
null
24
null
null
null
null
false
true
null
null
null
25
null
null
null
null
false
true
null
null
null
26
null
null
null
null
false
true
null
null
null
27
null
null
null
null
false
true
null
null
null
28
null
null
null
null
false
true
null
null
null
29
null
null
null
null
false
true
null
null
null
30
null
null
null
null
false
true
null
null
null
31
null
null
null
null
false
true
null
null
null
32
null
null
null
null
false
true
null
null
null
33
null
null
null
null
false
true
null
null
null
34
null
null
null
null
false
true
null
null
null
35
null
null
null
null
false
true
null
null
null
36
null
null
null
null
false
true
null
null
null
37
null
null
null
null
false
true
null
null
null
38
null
null
null
null
false
true
null
null
null
39
null
null
null
null
false
true
null
null
null
40
null
null
null
null
false
true
null
null
null
41
null
null
null
null
false
true
null
null
null
42
null
null
null
null
false
true
null
null
null
43
null
null
null
null
false
true
null
null
null
44
null
null
null
null
false
true
null
null
null
45
null
null
null
null
false
true
null
null
null
46
null
null
null
null
false
true
null
null
null
47
null
null
null
null
false
true
null
null
null
48
null
null
null
null
false
true
null
null
null
49
null
null
null
null
false
true
null
null
null
50
null
null
null
null
false
true
null
null
null
51
null
null
null
null
false
true
null
null
null
52
null
null
null
null
false
true
null
null
null
53
null
null
null
null
false
true
null
null
null
54
null
null
null
null
false
true
null
null
null
55
null
null
null
null
false
true
null
null
null
56
null
null
null
null
false
true
null
null
null
57
562
449
9
8
true
true
null
1
no
58
563
449
10
8
true
true
null
1
no
59
563
449
12
9
true
true
null
1
no
60
562
449
16
8
true
true
null
1
no
61
563
448
18
12
true
true
null
1
no
62
562
448
21
12
true
true
null
1
no
63
563
449
21
11
true
true
null
1
no
64
563
448
23
13
true
true
null
1
no
65
567
450
23
13
true
true
null
1
no
66
571
450
20
12
true
false
null
1
no
67
570
451
24
12
true
false
null
1
no
68
575
452
24
12
true
false
null
1
no
69
577
452
22
13
true
false
null
1
no
70
579
453
22
12
true
false
null
1
no
71
583
454
19
11
true
false
null
1
no
72
584
454
22
12
true
false
null
1
no
73
587
457
20
10
true
false
null
1
no
74
590
456
19
11
true
false
null
1
no
75
590
456
22
12
true
false
null
1
no
76
592
458
22
11
true
false
null
1
no
77
595
458
24
11
true
false
null
1
no
78
599
459
22
11
true
false
null
1
no
79
600
460
23
11
true
false
null
1
no
80
602
460
24
13
true
false
null
1
no
81
607
461
23
12
true
false
null
1
no
82
609
461
21
12
true
false
null
1
no
83
611
462
22
13
true
false
null
1
no
84
614
463
22
13
true
false
null
1
no
85
615
462
24
14
true
false
null
1
no
86
619
463
22
14
true
false
null
1
no
87
619
464
24
13
true
false
null
1
no
88
622
465
22
13
true
false
null
1
no
89
624
466
23
13
true
false
null
1
no
90
626
466
23
13
true
false
null
1
no
91
629
467
23
12
true
false
null
1
no
92
632
468
21
13
true
false
null
1
no
93
634
469
23
12
true
false
null
1
no
94
637
470
21
10
true
false
null
1
no
95
639
471
22
10
true
false
null
1
no
96
642
471
24
11
true
false
null
1
no
97
643
471
22
12
true
false
null
1
no
98
646
473
22
11
true
false
null
1
no
99
649
473
21
12
true
false
null
1
no
100
651
473
21
12
true
false
null
1
no
End of preview.

MVOT: MatrixCity Video Object Tracking

MVOT is a synthetic benchmark for video object tracking in satellite imagery. It contains short urban-scene videos rendered from MatrixCity, with frame-level target bounding boxes across controlled illumination, camera-tilt, trajectory, and occlusion conditions.

The dataset accompanies SatSAM2: Motion-Constrained Video Object Tracking in Satellite Imagery using Promptable SAM2 and Kalman Priors.

Dataset overview

Item Value
Video sequences 1,579
Regular sequences 1,487
Dedicated occlusion sequences 92
Total frames 157,900
Frames per sequence 100
Resolution 1024 x 1024
Frame rate 20 fps
Duration per sequence 5 seconds
Video codec H.264
Frames with a bounding box 151,006
Frames with a usable bounding box 145,902
Approximate download size 3.86 GiB

A usable bounding box is present and is not covered by an annotation-level invalid marker.

Conditions

Illumination

Lighting Sequences Frames
Day 942 94,200
Dusk 316 31,600
Night 321 32,100

Camera tilt

Camera tilt Sequences Frames
0 degrees 955 95,500
10 degrees 314 31,400
20 degrees 310 31,000

All dusk and night sequences use a camera tilt of 0 degrees.

Camera-trajectory region

Right, Left, Bottom, and Top identify four predefined camera-trajectory regions in MatrixCity. They do not describe the target's facing direction.

Region Sequences Frames
Bottom 354 35,400
Left 395 39,500
Right 372 37,200
Top 458 45,800

Occlusion subset

The dedicated occlusion subset contains 92 sequences and 9,200 frames. Bounding-box gaps form 109 contiguous events, with a mean event length of 23.65 frames.

Data generation

As described in the associated paper, MVOT was generated with MatrixCity rendering scripts in Unreal Engine. Camera positions were sampled at regular intervals along predefined trajectories at a fixed height. Scene pitch and illumination were varied under controlled settings, and target bounding boxes were obtained through the Unreal Engine actor-tracking API.

Directory structure

.
|-- README.md
|-- LICENSE
|-- CITATION.cff
|-- DATA_DICTIONARY.md
|-- annotations/
|   |-- regular/{lighting}/{camera_tilt}/{view}/*.csv
|   `-- occlusion/{lighting}/{camera_tilt}/{view}/*.csv
|-- data/
|   |-- metadata.jsonl
|   `-- videos/{subset}/{lighting}/{camera_tilt}/{view}/*.mp4
|-- manifests/
|   |-- checksums.sha256
|   |-- files.csv
|   `-- quality_report.json
`-- scripts/
    |-- upload_to_hub.py
    `-- validate_dataset.py

Loading the dataset

Install Hugging Face Datasets with video support, then load the repository:

from datasets import load_dataset

dataset = load_dataset(
    "Frank0666/MVOT",
    data_dir="data",
    split="train",
)

sample = dataset[0]
print(sample["sample_id"])
print(sample["annotations"][0])

For a local checkout:

from datasets import load_dataset

dataset = load_dataset("videofolder", data_dir="data", split="train")

The repository provides one default split. For model evaluation, define splits at the trajectory level where possible so that closely related clips do not appear in both training and evaluation sets.

Annotation format

Bounding boxes use pixel-space [x, y, width, height], where (x, y) is the top-left corner and the image origin is also at the top left. Each video record contains 100 ordered frame annotations. A missing box is represented by bbox_xywh: null and has_bbox: false.

Two frame-level fields support annotation filtering:

  • has_bbox indicates that all four box coordinates are present.
  • source_marked_invalid indicates that an invalid marker or range covers the frame.

For conservative training data selection, require has_bbox == true and source_marked_invalid == false.

The original frame-level occlusion labels are retained in occlusion_raw. Sequence membership in the dedicated occlusion subset is represented separately by subset == "occlusion".

See DATA_DICTIONARY.md for the complete metadata and annotation schema.

Validation

Run the structural validator from the repository root:

python scripts/validate_dataset.py
python scripts/validate_dataset.py --checksums

manifests/files.csv lists every video, annotation file, and metadata file in the release payload. manifests/checksums.sha256 provides their SHA-256 hashes.

Intended uses

MVOT supports single-object tracking, small-object localization, controlled robustness studies, and synthetic-to-real research across illumination, viewpoint, trajectory, and occlusion conditions.

Limitations

  • Target category names are not encoded in the annotations.
  • Synthetic scenes do not capture the full sensor, atmospheric, and viewing variability of real satellite imagery.
  • The repository does not provide an official train, validation, and test assignment.

MVOT is intended as a controlled complement to real-world datasets, not as a replacement for them.

License

MVOT is released under the MIT License. See LICENSE for the full terms.

Citation

@article{fan2025satsam2,
  title   = {SatSAM2: Motion-Constrained Video Object Tracking in Satellite Imagery using Promptable SAM2 and Kalman Priors},
  author  = {Fan, Ruijie and Ye, Junyan and Chen, Huan and Huang, Zilong and Wang, Xiaolei and Li, Weijia},
  journal = {arXiv preprint arXiv:2511.18264},
  year    = {2025},
  doi     = {10.48550/arXiv.2511.18264}
}
Downloads last month
-

Paper for Frank0666/MVOT