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 because of a cast error
Error code:   DatasetGenerationCastError
Exception:    DatasetGenerationCastError
Message:      An error occurred while generating the dataset

All the data files must have the same columns, but at some point there are 6 new columns ({'seed', 'kind', 'fold_count', 'histories', 'media_root', 'schema_version'}) and 2 missing columns ({'dataset', 'samples'}).

This happened while the json dataset builder was generating data using

hf://datasets/HaofanCao/deja-cue-data/data/ms_vost/manifest.json (at revision 5cdf053b6b4f6d9a9921c6878ed304442e22bde1), ['hf://datasets/HaofanCao/deja-cue-data@5cdf053b6b4f6d9a9921c6878ed304442e22bde1/data/ms_vost/annotations.json', 'hf://datasets/HaofanCao/deja-cue-data@5cdf053b6b4f6d9a9921c6878ed304442e22bde1/data/ms_vost/manifest.json', 'hf://datasets/HaofanCao/deja-cue-data@5cdf053b6b4f6d9a9921c6878ed304442e22bde1/data/ms_vost/prompts.json'], ['hf://datasets/HaofanCao/deja-cue-data@5cdf053b6b4f6d9a9921c6878ed304442e22bde1/data/ms_vost/annotations.json', 'hf://datasets/HaofanCao/deja-cue-data@5cdf053b6b4f6d9a9921c6878ed304442e22bde1/data/ms_vost/manifest.json', 'hf://datasets/HaofanCao/deja-cue-data@5cdf053b6b4f6d9a9921c6878ed304442e22bde1/data/ms_vost/prompts.json']

Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1856, in _prepare_split_single
                  writer.write_table(table)
                  ~~~~~~~~~~~~~~~~~~^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 811, 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 822, in _write_table
                  pa_table = table_cast(pa_table, self._schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2394, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2322, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              schema_version: int64
              kind: string
              seed: int64
              fold_count: int64
              media_root: string
              histories: struct<history_id: string, sequence_id: string, source_component_id: string, fold: int64, target: st (... 255 chars omitted)
                child 0, history_id: string
                child 1, sequence_id: string
                child 2, source_component_id: string
                child 3, fold: int64
                child 4, target: string
                child 5, video: string
                child 6, timestamps: string
                child 7, num_frames: int64
                child 8, recurrence: string
                child 9, states: list<item: struct<state_id: string, descriptions: list<item: string>, references: list<item: list<it (... 66 chars omitted)
                    child 0, item: struct<state_id: string, descriptions: list<item: string>, references: list<item: list<item: int64>> (... 54 chars omitted)
                        child 0, state_id: string
                        child 1, descriptions: list<item: string>
                            child 0, item: string
                        child 2, references: list<item: list<item: int64>>
                            child 0, item: list<item: int64>
                                child 0, item: int64
                        child 3, plausible_references: list<item: list<item: int64>>
                            child 0, item: list<item: int64>
                                child 0, item: int64
              -- schema metadata --
              pandas: '{"index_columns": [], "column_indexes": [], "columns": [{"name":' + 830
              to
              {'dataset': Value('string'), 'samples': {'id': Value('string'), 'target': Value('string'), 'recurrence': Value('string'), 'states': List({'label': Value('string'), 'visual_definition': Value('string')}), 'occurrences': List({'state_label': Value('string'), 'label_kind': Value('string'), 'earliest_plausible_start': Value('string'), 'first_definite_start': Value('string'), 'last_definite_end': Value('string'), 'latest_plausible_end': Value('string')}), 'media': {'file': Value('string')}}}
              because column names don't match
              
              During handling of the above exception, another exception occurred:
              
              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 1702, 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 1858, in _prepare_split_single
                  raise DatasetGenerationCastError.from_cast_error(
                  ...<4 lines>...
                  )
              datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
              
              All the data files must have the same columns, but at some point there are 6 new columns ({'seed', 'kind', 'fold_count', 'histories', 'media_root', 'schema_version'}) and 2 missing columns ({'dataset', 'samples'}).
              
              This happened while the json dataset builder was generating data using
              
              hf://datasets/HaofanCao/deja-cue-data/data/ms_vost/manifest.json (at revision 5cdf053b6b4f6d9a9921c6878ed304442e22bde1), ['hf://datasets/HaofanCao/deja-cue-data@5cdf053b6b4f6d9a9921c6878ed304442e22bde1/data/ms_vost/annotations.json', 'hf://datasets/HaofanCao/deja-cue-data@5cdf053b6b4f6d9a9921c6878ed304442e22bde1/data/ms_vost/manifest.json', 'hf://datasets/HaofanCao/deja-cue-data@5cdf053b6b4f6d9a9921c6878ed304442e22bde1/data/ms_vost/prompts.json'], ['hf://datasets/HaofanCao/deja-cue-data@5cdf053b6b4f6d9a9921c6878ed304442e22bde1/data/ms_vost/annotations.json', 'hf://datasets/HaofanCao/deja-cue-data@5cdf053b6b4f6d9a9921c6878ed304442e22bde1/data/ms_vost/manifest.json', 'hf://datasets/HaofanCao/deja-cue-data@5cdf053b6b4f6d9a9921c6878ed304442e22bde1/data/ms_vost/prompts.json']
              
              Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)

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.

dataset
string
samples
dict
MS-VOST
{ "id": "VOST-C0061847AD9130D55", "target": "Whole carrot segment being cut on a cutting board", "recurrence": "no", "states": [ { "label": "Intact single carrot unit", "visual_definition": "Whole-segment single carrot unit" }, { "label": "Half-segment carrot", "visual_defini...
MS-VOST
{ "id": "VOST-C01945484F3C1F847", "target": "Brown-skinned spherical object around the blue cutting board, with a white layered cross-section", "recurrence": "yes", "states": [ { "label": "Intact brown-skinned spherical object", "visual_definition": "The target has a closed spherical shape with ...
MS-VOST
{ "id": "VOST-C019E9A8CFD7978B7", "target": "Black continuous tubular film bag roll held in both hands, made of thin material with a smooth, stretchable surface", "recurrence": "no", "states": [ { "label": "Continuous tubular bag roll", "visual_definition": "The target is an end-to-end connected...
MS-VOST
{ "id": "VOST-C01DB15B4F988BD6C", "target": "Green-skinned, yellow-fleshed oval food ingredient on the wooden cutting board, with pale yellow flesh and greenish skin", "recurrence": "no", "states": [ { "label": "Intact single unit", "visual_definition": "The target is a single continuous oval en...
MS-VOST
{ "id": "VOST-C01E511A33416E2D5", "target": "Collection of fine white powdery material inside a white wide-mouthed bowl", "recurrence": "no", "states": [ { "label": "Dry powder pile", "visual_definition": "The material in the bowl is a loose white dry powder layer with dispersed particles, no we...
MS-VOST
{ "id": "VOST-C023AE4672126CD8D", "target": "Elongated wood-grained board on the miter saw worktable", "recurrence": "no", "states": [ { "label": "Intact long strip on worktable", "visual_definition": "The elongated board is intact and connected without a cut cross-section, placed as a single pi...
MS-VOST
{ "id": "VOST-C026116FB9B2E9A9E", "target": "Collection of milky-white liquid material in the container", "recurrence": "no", "states": [ { "label": "Concentrated retention in original liquid container", "visual_definition": "The white liquid material is stored concentrated inside the original l...
MS-VOST
{ "id": "VOST-C02DB6BC3B93EE23A", "target": "Dark paste-like pigment filler used for repairing holes in a wooden board", "recurrence": "no", "states": [ { "label": "Pigment stored concentrated in tube", "visual_definition": "The dark filler is entirely contained inside the pigment tube; the wood...
MS-VOST
{ "id": "VOST-C0325D931E4B58039", "target": "White raw dough blank to be processed", "recurrence": "no", "states": [ { "label": "Dough blank placed in roller gap of noodle-pressing machine", "visual_definition": "The dough lump is placed between the rollers of the noodle-pressing machine and has...
MS-VOST
{ "id": "VOST-C0364D58EE8839331", "target": "Green elongated melon/gourd to be cut", "recurrence": "no", "states": [ { "label": "Elongated melon/gourd cylinder intact and uncut", "visual_definition": "The elongated cylindrical melon/gourd material is placed inside the empty left bowl, maintainin...
MS-VOST
{ "id": "VOST-C0385DB16F8315B77", "target": "Fresh ginger piece to be chopped", "recurrence": "no", "states": [ { "label": "Whole ginger piece placed on cutting board, ready for initial cutting", "visual_definition": "The intact ginger piece is placed on the cutting board with the knife in posit...
MS-VOST
{ "id": "VOST-C041085E75601B072", "target": "Green long-stemmed vegetable to be cut and its cut round segments", "recurrence": "no", "states": [ { "label": "Whole bundle of long-stemmed vegetable secured on cutting board", "visual_definition": "Green long-stemmed vegetables are placed horizontal...
MS-VOST
{ "id": "VOST-C048486606F59DBDF", "target": "Two successive raw eggs to be cracked open, including eggshells and internal egg white and yolk; separated appearance of liquid egg and empty eggshell after cracking", "recurrence": "no", "states": [ { "label": "Intact uncracked raw egg", "visual_defi...
MS-VOST
{ "id": "VOST-C0486BE4C0CF4CE4E", "target": "A brown strip-shaped wooden handle, a part to be installed onto a small white cabinet, progressing from a loose and unsecured state through screw tightening to finally fit firmly against the cabinet surface", "recurrence": "no", "states": [ { "label": "Wood...
MS-VOST
{ "id": "VOST-C0606E9A2377CE0A7", "target": "The yellow butter block inside its packaging and the butter segments after cutting", "recurrence": "no", "states": [ { "label": "Butter wrapped inside tin foil", "visual_definition": "The yellow butter as a whole is covered by tin foil while both hand...
MS-VOST
{ "id": "VOST-C061C40B0E85D0A59", "target": "A silver cling film roll with a cardboard core, along with the cling film sheet separated and torn off from the roll", "recurrence": "no", "states": [ { "label": "Cling film roll held in hand, without pulling or tearing", "visual_definition": "The cyl...
MS-VOST
{ "id": "VOST-C06F0987FBF84A59B", "target": "The same piece of milky-white raw dough, tracked continuously throughout without replacing the dough body", "recurrence": "no", "states": [ { "label": "Dough held gathered together, not stretched", "visual_definition": "The dough is gathered in a ball...
MS-VOST
{ "id": "VOST-C06FDAECC1824C711", "target": "The same piece of milky-white raw dough, remaining the exact same dough subject throughout without replacement", "recurrence": "no", "states": [ { "label": "Dough resting in glass bowl, unprocessed", "visual_definition": "The dough rests stationary in...
MS-VOST
{ "id": "VOST-C07893654A241AD80", "target": "The same portion of brown wavy raw noodles, remaining the same food ingredient subject throughout without replacement", "recurrence": "no", "states": [ { "label": "Noodles resting in the plastic packaging box", "visual_definition": "The noodles remain...
MS-VOST
{ "id": "VOST-C07F04C00C3C80433", "target": "The wooden surface of the wooden fence / railing, which is the target object being painted", "recurrence": "no", "states": [ { "label": "Wooden fence unpainted, brush in the paint-dipping stage", "visual_definition": "The wooden fence surface remains ...
MS-VOST
{ "id": "VOST-C07FAD961BA1DA7A3", "target": "The same batch of white liquid batter with an unchanged main body; part remains in the original container and part is scooped out, transferred, and spread onto a frying pan, belonging to the same batter.", "recurrence": "yes", "states": [ { "label": "Batter...
MS-VOST
{ "id": "VOST-C0811868F3D7AB26C", "target": "A complete screw packaging kit (outer packaging box + screw bag + multiple sets of plastic protective caps), remaining the exact same set of materials throughout without change; it is simply unpacked step by step as the plastic caps are removed one by one.", "recurrenc...
MS-VOST
{ "id": "VOST-C08A92EB3090A1987", "target": "Two slices of whole-wheat toast as the same target object; the base bread remains unchanged, and the change consists only of jam being gradually added and spread to cover the surface.", "recurrence": "no", "states": [ { "label": "Two slices of toast with bu...
MS-VOST
{ "id": "VOST-C08B53B55D6588774", "target": "The same dark purple cloth / rag, remaining the same target object throughout without replacement", "recurrence": "no", "states": [ { "label": "Cloth held in hand above the water", "visual_definition": "The dark purple cloth is out of the water and he...
MS-VOST
{ "id": "VOST-CFFBC422B8C297AD7", "target": "The same segment of metal pipe fitting, remaining the exact same target throughout without replacement; changes arise from heating, sparks, and the red glow caused by rising temperature.", "recurrence": "no", "states": [ { "label": "Metal pipe fitting at ro...
MS-VOST
{ "id": "VOST-CFF8CE03C1125A77C", "target": "The tracked object is the meat-and-asparagus combination being processed, with 2 sets processed one after another; the asparagus is a green long-stemmed ingredient, the meat slice is red fresh raw meat, and the combination forms meat-wrapped asparagus.", "recurrence": ...
MS-VOST
{ "id": "VOST-CFF6E67DC7A692F05", "target": "The same rectangular wooden block, remaining the same target throughout without replacement; changes come only from orientation, working-face switching, and surface sanding.", "recurrence": "yes", "states": [ { "label": "Wooden block positioned and secured,...
MS-VOST
{ "id": "VOST-CFF289754D1A185DA", "target": "The same white translucent plastic film bag, remaining the exact same target throughout without replacement; changes arise from position, wrinkle morphology, being pushed in for storage, and being pressed and tidied.", "recurrence": "no", "states": [ { "lab...
MS-VOST
{ "id": "VOST-CFF1299533C4494D2", "target": "The same batch of light-colored soft dough, remaining the same main target throughout without replacement; changes consist only of morphological alterations after being taken out, divided into pieces, rubbed, patted, and pinched/pressed.", "recurrence": "yes", "state...
MS-VOST
{ "id": "VOST-CFE8E777BB172BCA8", "target": "The initial target is a whole roll of kitchen paper; after a single sheet of paper towel is separated, the target shifts to that paper towel sheet and the item wrapped inside it, remaining a continuous target under the same task chain throughout without replacement; chan...
MS-VOST
{ "id": "VOST-CFE17661B8744B40E", "target": "The tracked object is the same single onion, divided into two halves that are peeled one after another; the main onion flesh is the continuously tracked target while the peeled-off outer skins are waste rather than the main target; there is no target replacement througho...
MS-VOST
{ "id": "VOST-CFCD3623344716D4E", "target": "The tracked object is the same piece of onion flesh; the peeled-off outer skin and pointed-corner waste are byproducts, whereas the main target is always the onion body itself without target replacement throughout.", "recurrence": "no", "states": [ { "label...
MS-VOST
{ "id": "VOST-CFAE261C6171F19C5", "target": "Batter mixture", "recurrence": "no", "states": [ { "label": "Small bowl of batter resting stationary, scooping not yet started", "visual_definition": "Small bowl of batter resting stationary, scooping not yet started" }, { "label": "Usin...
MS-VOST
{ "id": "VOST-CFABD753B11627C5D", "target": "Mud ball", "recurrence": "no", "states": [ { "label": "Loose soil state", "visual_definition": "The soil lies loosely spread on the ground without any formed clumps" }, { "label": "Kneaded and shaped soil ball", "visual_definition"...
MS-VOST
{ "id": "VOST-CFAA166849B98837D", "target": "Light beige dough dusted with white dry flour on the surface, placed on a wooden kneading board;", "recurrence": "yes", "states": [ { "label": "Flour dough resting on the pastry board, unkneaded", "visual_definition": "The flour dough is on the kneadi...
MS-VOST
{ "id": "VOST-CFA2EC521975B5DF9", "target": "Garlic clove on a white cutting board", "recurrence": "no", "states": [ { "label": "Garlic clove resting on cutting board, not yet picked up for peeling", "visual_definition": "The pale yellow garlic clove rests stationary on the white square cutting ...
MS-VOST
{ "id": "VOST-CF8557E2A673E9D05", "target": "Onion", "recurrence": "no", "states": [ { "label": "Onion resting on the cutting board, uncut", "visual_definition": "The onion piece rests on the wooden cutting board surface before the knife cuts into the onion, maintaining its original block form" ...
MS-VOST
{ "id": "VOST-CF6B43724B0E51A02", "target": "Vegetables to be cut", "recurrence": "no", "states": [ { "label": "Vegetable resting in plate, awaiting use", "visual_definition": "The elongated vegetable rests stationary inside the empty metal plate without being picked up and with no cutting opera...
MS-VOST
{ "id": "VOST-CF6075ECB7A01F7F5", "target": "Pumpkin", "recurrence": "no", "states": [ { "label": "Pumpkin resting on the cutting board, awaiting operation", "visual_definition": "The pumpkin piece rests stationary on the white cutting board before the knife cuts downward, waiting to be cut" ...
MS-VOST
{ "id": "VOST-CF5BBA915D4EF3BF4", "target": "Chunk material and peel waste inside the container.", "recurrence": "no", "states": [ { "label": "S1", "visual_definition": "Material stored inside the container, resting stationary awaiting pouring" }, { "label": "Container tilted, star...
MS-VOST
{ "id": "VOST-CF552737DB8BAE96E", "target": "White root vegetable", "recurrence": "no", "states": [ { "label": "Round vegetable slices resting on cutting board, awaiting cutting", "visual_definition": "Round vegetable slices resting on cutting board, awaiting cutting" }, { "label":...
MS-VOST
{ "id": "VOST-CF43E20EC7C30293B", "target": "Garlic clove", "recurrence": "no", "states": [ { "label": "Garlic clove resting on tabletop, awaiting processing", "visual_definition": "The garlic clove rests stationary on the wooden tabletop before peeling and cutting operations begin" }, {...
MS-VOST
{ "id": "VOST-CF438B7B2E452C73C", "target": "Walnut", "recurrence": "no", "states": [ { "label": "Walnut resting inside basket, awaiting picking up and cracking open", "visual_definition": "All walnuts are stored in the left woven basket without being picked up and with no cracking operation" ...
MS-VOST
{ "id": "VOST-CF3FD2F992EA3A125", "target": "Zucchini", "recurrence": "no", "states": [ { "label": "Zucchini resting on cutting board, awaiting peeling", "visual_definition": "The zucchini rests on the green cutting board with its outer skin intact before the peeling action begins" }, { ...
MS-VOST
{ "id": "VOST-CF3B19D118BCE6504", "target": "Onion", "recurrence": "no", "states": [ { "label": "Onion resting on cutting board, awaiting preparation", "visual_definition": "The intact onion rests on the cutting board before top-and-bottom trimming, halving, and peeling operations are performed"...
MS-VOST
{ "id": "VOST-CF3967C914FCAFD76", "target": "Raw chicken breast meat", "recurrence": "no", "states": [ { "label": "Chicken breast meat sealed inside the packaging box, awaiting unpacking", "visual_definition": "The chicken breast meat remains sealed inside the original green packaging box, the p...
MS-VOST
{ "id": "VOST-CF2430ADE476BA738", "target": "Paper picture", "recurrence": "no", "states": [ { "label": "Paper picture lying flat on the tabletop, awaiting cutting", "visual_definition": "The picture rests stationary on the wooden tabletop, not yet picked up and not yet cut with scissors" },...
MS-VOST
{ "id": "VOST-CF23EDEC9A5902ADF", "target": "White coated area on the wooden board surface", "recurrence": "no", "states": [ { "label": "Wooden board awaiting painting, white coating not yet formed", "visual_definition": "The wooden board lies flat on the work mat, the target area is not covered...
MS-VOST
{ "id": "VOST-CF1CB44546EC84732", "target": "Metal brake fittings (sheet-metal ring parts)", "recurrence": "no", "states": [ { "label": "Metal fittings sealed inside plastic bag, awaiting unsealing", "visual_definition": "The fittings are sealed inside the plastic bag and have not been taken out...
MS-VOST
{ "id": "VOST-CF04298E04C513B39", "target": "Raw dough flatbreads on the baking pan", "recurrence": "no", "states": [ { "label": "Raw dough flatbreads resting on baking pan, not yet spread with tomato sauce", "visual_definition": "The flatbreads are arranged on the baking pan with no red tomato ...
MS-VOST
{ "id": "VOST-CF03AECF8AA7F6271", "target": "Carrot", "recurrence": "no", "states": [ { "label": "Carrot section resting stationary on cutting board, cutting not yet started", "visual_definition": "The carrot section is placed on the cutting board, remaining intact and awaiting downward cutting"...
MS-VOST
{ "id": "VOST-CEF85D3FBAFE421E7", "target": "Wooden board", "recurrence": "no", "states": [ { "label": "Wooden board secured on workbench, awaiting planing processing", "visual_definition": "The wooden board is secured on the workbench by a clamp, and its surface has not yet been processed by th...
MS-VOST
{ "id": "VOST-CEEC46E5179ED922C", "target": "Bok choy / Chinese cabbage", "recurrence": "no", "states": [ { "label": "Bok choy resting on cutting board, cutting not yet started", "visual_definition": "Bok choy is placed on a wooden cutting board, a hand presses down on the leaves, and a knife co...
MS-VOST
{ "id": "VOST-CEEAFADF30DEC1240", "target": "Baking parchment paper", "recurrence": "no", "states": [ { "label": "Parchment paper laid flat on tabletop, awaiting cutting", "visual_definition": "The parchment paper lies unfolded on the tabletop before scissor cutting begins" }, { "l...
MS-VOST
{ "id": "VOST-CEEA732AB7BD33262", "target": "Onion", "recurrence": "no", "states": [ { "label": "Onions with outer skin on, awaiting peeling", "visual_definition": "The onions remain intact and peeling of the outer skin has not yet begun" }, { "label": "Peeling the outer skin of th...
MS-VOST
{ "id": "VOST-CEE4A7761593E59F5", "target": "Patterned area on the paper", "recurrence": "yes", "states": [ { "label": "Paper with pattern laid flat, awaiting coloring", "visual_definition": "The white paper and pattern are clearly visible before coloring begins" }, { "label": "Ini...
MS-VOST
{ "id": "VOST-CEDF153F9309FB038", "target": "Orange food ingredient (carrot slices)", "recurrence": "no", "states": [ { "label": "Round carrot slices placed on cutting board, awaiting cutting", "visual_definition": "The carrot slices are intact before the knife cuts downward" }, { ...
MS-VOST
{ "id": "VOST-CEDE42D55896A4594", "target": "Potato", "recurrence": "no", "states": [ { "label": "Potatoes placed on cutting board, awaiting cutting", "visual_definition": "The potatoes remain intact before downward cutting begins" }, { "label": "Initiating cutting and beginning to...
MS-VOST
{ "id": "VOST-CED60FE8FD38723FE", "target": "Ginger", "recurrence": "no", "states": [ { "label": "Ginger placed on cutting board, awaiting processing", "visual_definition": "The ginger remains intact and peeling has not yet begun" }, { "label": "Ginger peeling stage", "visual...
MS-VOST
{ "id": "VOST-CECEE241AB9136E34", "target": "Peach", "recurrence": "no", "states": [ { "label": "Peach held in hand awaiting cutting", "visual_definition": "The peeled peach is held intact in hand before the knife cuts into the flesh to begin taking pieces" }, { "label": "Initiatin...
MS-VOST
{ "id": "VOST-CECBB71923B9B2FED", "target": "Sausage", "recurrence": "no", "states": [ { "label": "Sausage sealed inside bag awaiting removal", "visual_definition": "The sausage is still inside the plastic bag and has not yet been taken out and placed on the cutting board" }, { "la...
MS-VOST
{ "id": "VOST-CEAF7230DB479C62D", "target": "Minced meat filling", "recurrence": "no", "states": [ { "label": "Whole block of minced meat held in hand awaiting splitting", "visual_definition": "The minced meat is a single whole block held in both hands before tearing and dividing begin" }, ...
MS-VOST
{ "id": "VOST-CE9B4571FC4E88E7B", "target": "Long green scallion stalks on the cutting board", "recurrence": "no", "states": [ { "label": "Long green scallions with intact shape and no chopping marks, resting stationary on the cutting board", "visual_definition": "The object maintains an intact ...
MS-VOST
{ "id": "VOST-CE8A00AF99784F829", "target": "White dough in the mixing bowl", "recurrence": "no", "states": [ { "label": "Dough resting stationary in the bowl awaiting kneading", "visual_definition": "The dough is placed smoothly inside the bowl with no kneading marks on its surface and has not ...
MS-VOST
{ "id": "VOST-CE7A265F7325DAF04", "target": "Cilantro / coriander (green leafy vegetable)", "recurrence": "no", "states": [ { "label": "Gathering the vegetable together", "visual_definition": "Both hands gather and bring the cilantro together into a bundle, before pressing it down firmly onto th...
MS-VOST
{ "id": "VOST-CE79E51443118C4BA", "target": "Carrot", "recurrence": "no", "states": [ { "label": "Whole carrot awaiting cutting", "visual_definition": "The whole carrot is placed on the cutting board before downward cutting with the knife begins" }, { "label": "Continuously slicing...
MS-VOST
{ "id": "VOST-C08C2D1175BFAD3F6", "target": "Carrot", "recurrence": "no", "states": [ { "label": "Carrot placed on cutting board awaiting cutting", "visual_definition": "The whole carrot rests on the cutting board before halving and division begin" }, { "label": "Continuously divid...
MS-VOST
{ "id": "VOST-C09108E7BCE20244E", "target": "Chicken breast", "recurrence": "no", "states": [ { "label": "Chicken breast meat placed on cutting board awaiting cutting", "visual_definition": "The whole piece of chicken breast meat rests on the cutting board before downward knife cutting begins" ...
MS-VOST
{ "id": "VOST-C0A444B55C148EAA7", "target": "Mango flesh", "recurrence": "no", "states": [ { "label": "Mango flesh positioned awaiting cutting", "visual_definition": "The mango flesh piece rests on the cutting board held steady by hand before the knife cuts downward" }, { "label": ...
MS-VOST
{ "id": "VOST-C0A2FFC90C3566D9E", "target": "Food ingredient inside the food storage box", "recurrence": "no", "states": [ { "label": "Food ingredient inside the storage box awaiting lid opening", "visual_definition": "The storage box is held in hand inside the refrigerator with the lid closed, ...
MS-VOST
{ "id": "VOST-C0995A5DE03FF02F4", "target": "Tomato", "recurrence": "no", "states": [ { "label": "Tomato awaiting cutting", "visual_definition": "The tomato is placed on the cutting board and held in position by hand before the knife cuts downward" }, { "label": "Cutting the tomato...
MS-VOST
{ "id": "VOST-C0984179E1A049B6C", "target": "Carrot", "recurrence": "no", "states": [ { "label": "Carrot awaiting peeling", "visual_definition": "The carrot is held in hand, and the peeling operation has not yet been initiated" }, { "label": "Continuously peeling the carrot", ...
MS-VOST
{ "id": "VOST-C0979845935F37F3F", "target": "Cucumber", "recurrence": "no", "states": [ { "label": "Cucumber placed on cutting board awaiting cutting", "visual_definition": "The cucumber rests on the cutting board before downward knife cutting begins" }, { "label": "Continuously cu...
MS-VOST
{ "id": "VOST-C0B5054AB897C05B2", "target": "Coloring area on the coloring paper", "recurrence": "no", "states": [ { "label": "Coloring sheet awaiting coloring", "visual_definition": "The coloring sheet lies flat on the desk with local parts of the pattern not yet colored in and the colored pen/...
MS-VOST
{ "id": "VOST-C0C5C6BF110D2F1CC", "target": "Garlic", "recurrence": "no", "states": [ { "label": "Garlic placed on the cutting board awaiting cutting", "visual_definition": "The garlic is placed on the cutting board and held in position by hand, with no downward knife cuts made yet" }, {...
MS-VOST
{ "id": "VOST-C0BA58F259784D812", "target": "Carrot", "recurrence": "no", "states": [ { "label": "Carrot placed on cutting board awaiting cutting", "visual_definition": "The carrot is placed on the cutting board and held in position by hand, with no downward knife cuts made yet" }, { ...
MS-VOST
{ "id": "VOST-C0CD1DDF37C34D89A", "target": "Tomato", "recurrence": "no", "states": [ { "label": "Tomato placed on cutting board awaiting cutting", "visual_definition": "The tomato rests on the cutting board held in position by hand before the knife cuts downward" }, { "label": "In...
MS-VOST
{ "id": "VOST-C0DC2835CA57BA948", "target": "Potato", "recurrence": "no", "states": [ { "label": "Potato held in hand awaiting peeling", "visual_definition": "The potato is held in hand before the peeler begins contacting the potato skin" }, { "label": "Initiating peeling", "...
MS-VOST
{ "id": "VOST-C10EC0465F185A688", "target": "Car floor mat", "recurrence": "no", "states": [ { "label": "Floor mat placed inside the trunk awaiting removal", "visual_definition": "The floor mat is stored in the trunk and not yet taken out, while only dust inside the trunk is being cleaned" }...
MS-VOST
{ "id": "VOST-C107F3821CA911D41", "target": "Mud / clay soil", "recurrence": "no", "states": [ { "label": "Mud taken out from the mud pile awaiting processing", "visual_definition": "The mud separates from the original mud pile and is picked up in hand, not yet entering the kneading process" ...
MS-VOST
{ "id": "VOST-C0FCF23C40ADA2531", "target": "Garlic clove", "recurrence": "yes", "states": [ { "label": "Garlic clove resting stationary awaiting processing", "visual_definition": "The garlic clove maintains its original form, and cutting, smashing, and peeling operations have not started" }...
MS-VOST
{ "id": "VOST-C0FB9367198C96056", "target": "Banana", "recurrence": "no", "states": [ { "label": "Intact banana resting stationary unpeeled", "visual_definition": "The banana peel is intact, with no peeling or cutting action, maintaining its whole form" }, { "label": "Banana peelin...
MS-VOST
{ "id": "VOST-C0F27263DEAEFE6F4", "target": "Deck / set of playing cards", "recurrence": "yes", "states": [ { "label": "Closed and stacked playing cards", "visual_definition": "The playing cards are stacked and gathered together into a compact overlapping contour overall, with no large area of c...
MS-VOST
{ "id": "VOST-C0EDBEF668DDAF705", "target": "Kimchi filling", "recurrence": "no", "states": [ { "label": "Kimchi filling contained inside the container awaiting use", "visual_definition": "The filling is inside the container before the spoon begins scooping the filling" }, { "label...
MS-VOST
{ "id": "VOST-C0DF0D8FD193893F3", "target": "Flour / flour dough", "recurrence": "no", "states": [ { "label": "Flour sealed inside the container awaiting processing", "visual_definition": "The flour is held in a lidded container, the lid has not been lifted open, and the flour has not been taken...
End of preview.

Déjà Cue Fixed-Feature Evaluation Assets

Frozen features, fixed evaluation settings, exact reference windows, and statistical results for Déjà Cue.

Haofan Cao · Zhichao You · Yunkai Yang · Liang Guo · Jie Wang · Chongshou Li

arXiv Companion GitHub code Data terms

Overview of the Déjà Cue retrieval framework

📄 Abstract

Tracking links observations of the same object through visual change, yet cannot by itself determine when the object is empty or filled, intact or cut. Déjà Cue formulates identity-conditioned state-moment retrieval: given a tracked-object history and alternative state descriptions, it localizes an interval in which each described state holds. The method turns the alternatives into a vocabulary-relative coordinate system, subtracts their state-balanced centroid from each description, calibrates frame scores, and scans multiple durations within contiguous visible runs using a frozen encoder. On 78 VOST histories, changing only the query reference nearly doubles R@1 at tIoU 0.5 from 10.3% to 20.5% and raises Top-1 tIoU from 16.0% to 21.5%.

🔥 News

  • 2026.08.05 · The official code and fixed-feature evaluation assets are now open source.

📦 Dataset Summary

This repository contains the same data/ directory as the companion code release. It reproduces retrieval, interval selection, metrics, and statistical calculations without redistributing native video or rerunning the frozen encoder.

Dataset page: HaofanCao/deja-cue-data

Group Contents
VOST evaluation 78 histories, 156 states, 312 descriptions, reference intervals, and frozen features
Prompt variants Raw, photo, definite, and normalized three-form text features
Development set 5 histories, 13 states, 26 descriptions, 46 positive episodes, and 86 training records
Seven-history evaluation 7 histories, 16 states, 32 descriptions, 59 reference episodes, and 8 auxiliary tracks
Reference results Exact selected windows and the rows used for the reported statistical calculations
Evaluation settings Fixed VOST split lists, final cohort information, method settings, and checksums

The dataset contains 288 files under data/. DATA_MANIFEST.json records the byte size and SHA-256 digest of every one.

The Hub's n<1K label describes the number of dataset examples, not storage size. The complete repository occupies about 51.5 MB.

⬇️ Download and Verify

The complete dataset is published at HaofanCao/deja-cue-data. With the current Hugging Face CLI, download it with:

export DEJA_CUE_DATASET_ID=HaofanCao/deja-cue-data
hf download "$DEJA_CUE_DATASET_ID" --type dataset --local-dir deja-cue-data
cd deja-cue-data
python verify_data.py

The script checks every file name, byte size, and SHA-256 value. A Python download is equivalent:

from pathlib import Path

from huggingface_hub import snapshot_download

dataset_root = Path(
    snapshot_download(
        repo_id="HaofanCao/deja-cue-data",
        repo_type="dataset",
        local_dir="deja-cue-data",
    )
)

🔎 Load One History

import numpy as np

visual_path = dataset_root / "data/features/siglip2/H001/visual_features.npz"
with np.load(visual_path, allow_pickle=False) as arrays:
    frame_indices = arrays["frame_indices"]
    visual_features = arrays["visual_features"]
    visibility_count = arrays["visibility_count"]

print(frame_indices.shape, visual_features.shape, visibility_count.shape)

Primary visual archives contain:

Field Meaning
frame_indices Original temporal indices for observed target frames
visual_features Unit-normalized 768-dimensional SigLIP 2 features
visibility_count Number of target-lineage masks contributing at each frame

Primary text archives contain state_ids, state_texts, and unit-normalized text_features. Prompt-variant archives additionally contain variant_names. JSON manifests define sibling-state grouping, inclusive reference intervals, source-component aggregation, and relative feature paths.

🔗 Use with the Code Release

The companion GitHub repository already includes the same data/ files, so a complete clone can reproduce the CPU results immediately:

git clone https://github.com/HaofanCao/DejaCue.git
cd DejaCue
python -m pip install -e ".[test]"
python scripts/reproduce_main.py --device cpu

When using a separate Hub download, place or link its data/ directory at the code repository root without changing the relative paths. The code repository supplies the data loaders, retrieval implementation, reproduction commands, annotation and feature tools, and tests.

🗂️ Repository Layout

data/
  benchmark.json              Primary benchmark manifest
  features/                   VOST visual, text, and prompt features
  learned/development/        Frozen learned-decoder development inputs
  seven_history/              Seven-history features and auxiliary tracks
  reference/                  Exact windows and statistical results
  protocol/vost/              Fixed VOST split lists and protocol notes
DATA_MANIFEST.json            File sizes and SHA-256 checksums
DATA_LICENSE.md               Data and derived-feature terms
verify_data.py                File verification script

Evaluation identifiers such as H001 are stable IDs used to align features, states, descriptions, and references. All manifest paths are relative to the repository root.

🌐 Data Sources

Asset family Upstream research source What this repository contains
Primary histories VOST Derived object-local features, evaluation IDs, fixed settings, and reference intervals
Six seven-history sequences HyperNeRF Derived target and auxiliary-track features plus evaluation metadata
Coffee-martini sequence Neural 3D Video Derived target features and evaluation metadata
Recurrence diagnostic D-NeRF Predictions and reference intervals, not rendered RGB
Feature encoder SigLIP 2 Unit-normalized derived embeddings; no model weights

Feature extraction uses google/siglip2-base-patch16-224 at revision 75de2d55ec2d0b4efc50b3e9ad70dba96a7b2fa2. Upstream media and model terms continue to apply to derived assets; see DATA_LICENSE.md.

📝 Annotation and Included Files

For VOST, two state descriptions per state are fixed before labeling. Two annotators independently label target-lineage frames as pre-state, transition, post-state, or unobserved while blind to method scores, and every disagreement is resolved to consensus. The 100-history selection retains 78 qualifying histories and excludes 22 with no qualifying event.

The dataset includes the split lists, final cohort information, annotation validation scripts, and agreement and annotation-sensitivity summaries. It does not redistribute native RGB, target-lineage masks, pretrained encoder weights, generated decoder checkpoints, the two complete annotation files, or the resolved framewise consensus.

✅ Intended Uses

  • Reproduction of the primary and extended VOST evaluations.
  • Reproduction of the seven-history coordinate study.
  • Training the documented eight learned decoders on the five-history development set and evaluating generated checkpoints.
  • Recalculation of the reported robustness, diagnostic, and statistical summaries.
  • Comparison of alternative retrieval methods under the same frozen representation and evaluation protocol.

⚠️ Limitations

The task assumes target-lineage masks, a closed sibling vocabulary, and one selected interval per description. It has no calibrated null output for an absent state. VOST uses binary event-adjacent states and scores one designated occurrence; the dataset also contains 13 additional pre-state and 20 additional post-state occurrences. The seven-history set adds multi-state, interior, recurrent, and progressive cases, but remains small and its three independently held-out histories show mixed coordinate effects.

Fixed features isolate retrieval logic from encoder execution and do not establish behavior for other encoders. Tracking stresses use controlled feature mixing and identity switches rather than pixel-level mask drift. The release does not establish performance for automatically discovered identities, open vocabularies, broad multi-state collections, every natural tracking failure, or hardware-specific latency.

🛡️ Privacy and Responsible Use

No native RGB, video, masks, direct identifiers, or model weights are included. The embeddings remain derived from external research media and must be handled under the corresponding source terms. These assets are intended for scientific retrieval and reproducibility research, not biometric identification, surveillance, or inferring sensitive attributes.

⚖️ Licensing

This repository combines metadata created for Déjà Cue with embeddings derived from external research data, so its Hugging Face license field is other. Read DATA_LICENSE.md before reuse or redistribution, and retain the upstream attributions required for the source assets.

📚 Citation

If you use these assets, please cite the official paper record:

@misc{cao2026dejacue,
  title         = {Déjà Cue: Localizing States in Object Histories via Vocabulary-Relative Coordinates},
  author        = {Haofan Cao and Zhichao You and Yunkai Yang and Liang Guo and Jie Wang and Chongshou Li},
  year          = {2026},
  eprint        = {2608.02044},
  archivePrefix = {arXiv},
  primaryClass  = {cs.CV},
  url           = {https://arxiv.org/abs/2608.02044}
}

Paper: arXiv:2608.02044

Downloads last month
194

Papers for HaofanCao/deja-cue-data