Datasets:

LiYacheng
/

r4d-bench-qa

Error code:   StreamingRowsError
Exception:    CastError
Message:      Couldn't cast
query_id: string
text_en: string
text_zh: string
target_track_ids: list<item: string>
  child 0, item: string
target_count: int64
requires_motion_understanding: bool
query_type: string
temporal_relation: string
expected_temporal_anchor: string
scene: string
note: string
distractor: string
cut_lemon_q6: list<item: string>
  child 0, item: string
espresso_q5: list<item: null>
  child 0, item: null
espresso_q7: list<item: string>
  child 0, item: string
keyboard_q5: list<item: null>
  child 0, item: null
cook-spinach_q4: list<item: string>
  child 0, item: string
flame_steak_q7: list<item: string>
  child 0, item: string
cut_lemon_q4: list<item: string>
  child 0, item: string
torchchocolate_q7: list<item: string>
  child 0, item: string
americano_q8: list<item: null>
  child 0, item: null
keyboard_q2: list<item: string>
  child 0, item: string
split_cookie_q4: list<item: string>
  child 0, item: string
sear_steak_q4: list<item: string>
  child 0, item: string
flame_salmon_q3: list<item: string>
  child 0, item: string
split_cookie_q7: list<item: string>
  child 0, item: string
cook-spinach_q5: list<item: null>
  child 0, item: null
sear_steak_q6: list<item: null>
  child 0, item: null
split_cookie_q3: list<item: string>
  child 0, item: string
espresso_q3: list<item: string>
  child 0, item: string
cut_lemon_q8: list<item: null>
  child 0, item: null
americano_q10: list<item: string>
  child 0, item: string
americano_q9: list<item: string>
  child 0, item: string
split_cookie_q
...
ak_q5: list<item: null>
  child 0, item: null
cut_lemon_q1: list<item: string>
  child 0, item: string
flame_salmon_q4: list<item: string>
  child 0, item: string
cut_lemon_q5: list<item: string>
  child 0, item: string
torchchocolate_q3: list<item: string>
  child 0, item: string
sear_steak_q5: list<item: null>
  child 0, item: null
cook-spinach_q2: list<item: string>
  child 0, item: string
americano_q5: list<item: string>
  child 0, item: string
cook-spinach_q3: list<item: string>
  child 0, item: string
espresso_q6: list<item: null>
  child 0, item: null
split_cookie_q2: list<item: string>
  child 0, item: string
coffee_martini_q6: list<item: null>
  child 0, item: null
keyboard_q1: list<item: string>
  child 0, item: string
cut_roasted_beef_q1: list<item: string>
  child 0, item: string
split_cookie_q5: list<item: null>
  child 0, item: null
split_cookie_q8: list<item: null>
  child 0, item: null
coffee_martini_q1: list<item: string>
  child 0, item: string
americano_q2: list<item: string>
  child 0, item: string
sear_steak_q1: list<item: string>
  child 0, item: string
americano_q3: list<item: string>
  child 0, item: string
cut_lemon_q2: list<item: string>
  child 0, item: string
cut_roasted_beef_q7: list<item: string>
  child 0, item: string
flame_salmon_q7: list<item: string>
  child 0, item: string
cook-spinach_q6: list<item: null>
  child 0, item: null
espresso_q4: list<item: string>
  child 0, item: string
flame_steak_q3: list<item: string>
  child 0, item: string
to
{'cut_lemon_q1': List(Value('string')), 'cut_lemon_q2': List(Value('string')), 'cut_lemon_q3': List(Value('string')), 'cut_lemon_q4': List(Value('string')), 'cut_lemon_q5': List(Value('string')), 'cut_lemon_q6': List(Value('string')), 'cut_lemon_q7': List(Value('null')), 'cut_lemon_q8': List(Value('null')), 'espresso_q1': List(Value('string')), 'espresso_q2': List(Value('string')), 'espresso_q3': List(Value('string')), 'espresso_q4': List(Value('string')), 'espresso_q5': List(Value('null')), 'espresso_q6': List(Value('null')), 'espresso_q7': List(Value('string')), 'espresso_q8': List(Value('string')), 'keyboard_q1': List(Value('string')), 'keyboard_q2': List(Value('string')), 'keyboard_q3': List(Value('string')), 'keyboard_q4': List(Value('null')), 'keyboard_q5': List(Value('null')), 'keyboard_q6': List(Value('string')), 'torchchocolate_q1': List(Value('string')), 'torchchocolate_q2': List(Value('string')), 'torchchocolate_q3': List(Value('string')), 'torchchocolate_q4': List(Value('null')), 'torchchocolate_q5': List(Value('null')), 'torchchocolate_q6': List(Value('string')), 'torchchocolate_q7': List(Value('string')), 'cook-spinach_q1': List(Value('string')), 'cook-spinach_q2': List(Value('string')), 'cook-spinach_q3': List(Value('string')), 'cook-spinach_q4': List(Value('string')), 'cook-spinach_q5': List(Value('null')), 'cook-spinach_q6': List(Value('null')), 'cook-spinach_q7': List(Value('string')), 'cut_roasted_beef_q1': List(Value('string')), 'cut_roasted_beef_q2': List
...
sear_steak_q3': List(Value('string')), 'sear_steak_q4': List(Value('string')), 'sear_steak_q5': List(Value('null')), 'sear_steak_q6': List(Value('null')), 'sear_steak_q7': List(Value('string')), 'split_cookie_q1': List(Value('string')), 'split_cookie_q2': List(Value('string')), 'split_cookie_q3': List(Value('string')), 'split_cookie_q4': List(Value('string')), 'split_cookie_q5': List(Value('null')), 'split_cookie_q6': List(Value('string')), 'split_cookie_q7': List(Value('string')), 'split_cookie_q8': List(Value('null')), 'americano_q1': List(Value('string')), 'americano_q2': List(Value('string')), 'americano_q3': List(Value('string')), 'americano_q4': List(Value('string')), 'americano_q5': List(Value('string')), 'americano_q6': List(Value('string')), 'americano_q7': List(Value('null')), 'americano_q8': List(Value('null')), 'americano_q9': List(Value('string')), 'americano_q10': List(Value('string')), 'coffee_martini_q1': List(Value('string')), 'coffee_martini_q2': List(Value('string')), 'coffee_martini_q3': List(Value('string')), 'coffee_martini_q4': List(Value('string')), 'coffee_martini_q5': List(Value('null')), 'coffee_martini_q6': List(Value('null')), 'coffee_martini_q7': List(Value('string')), 'flame_steak_q1': List(Value('string')), 'flame_steak_q2': List(Value('string')), 'flame_steak_q3': List(Value('string')), 'flame_steak_q4': List(Value('string')), 'flame_steak_q5': List(Value('null')), 'flame_steak_q6': List(Value('null')), 'flame_steak_q7': List(Value('string'))}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 99, in get_rows_or_raise
                  return get_rows(
                         ^^^^^^^^^
                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 77, in get_rows
                  rows_plus_one = list(itertools.islice(ds, rows_max_number + 1))
                                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2690, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2227, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2251, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 494, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 384, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 295, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 128, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2321, in table_cast
                  return cast_table_to_schema(table, schema)
                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2249, in cast_table_to_schema
                  raise CastError(
              datasets.table.CastError: Couldn't cast
              query_id: string
              text_en: string
              text_zh: string
              target_track_ids: list<item: string>
                child 0, item: string
              target_count: int64
              requires_motion_understanding: bool
              query_type: string
              temporal_relation: string
              expected_temporal_anchor: string
              scene: string
              note: string
              distractor: string
              cut_lemon_q6: list<item: string>
                child 0, item: string
              espresso_q5: list<item: null>
                child 0, item: null
              espresso_q7: list<item: string>
                child 0, item: string
              keyboard_q5: list<item: null>
                child 0, item: null
              cook-spinach_q4: list<item: string>
                child 0, item: string
              flame_steak_q7: list<item: string>
                child 0, item: string
              cut_lemon_q4: list<item: string>
                child 0, item: string
              torchchocolate_q7: list<item: string>
                child 0, item: string
              americano_q8: list<item: null>
                child 0, item: null
              keyboard_q2: list<item: string>
                child 0, item: string
              split_cookie_q4: list<item: string>
                child 0, item: string
              sear_steak_q4: list<item: string>
                child 0, item: string
              flame_salmon_q3: list<item: string>
                child 0, item: string
              split_cookie_q7: list<item: string>
                child 0, item: string
              cook-spinach_q5: list<item: null>
                child 0, item: null
              sear_steak_q6: list<item: null>
                child 0, item: null
              split_cookie_q3: list<item: string>
                child 0, item: string
              espresso_q3: list<item: string>
                child 0, item: string
              cut_lemon_q8: list<item: null>
                child 0, item: null
              americano_q10: list<item: string>
                child 0, item: string
              americano_q9: list<item: string>
                child 0, item: string
              split_cookie_q
              ...
              ak_q5: list<item: null>
                child 0, item: null
              cut_lemon_q1: list<item: string>
                child 0, item: string
              flame_salmon_q4: list<item: string>
                child 0, item: string
              cut_lemon_q5: list<item: string>
                child 0, item: string
              torchchocolate_q3: list<item: string>
                child 0, item: string
              sear_steak_q5: list<item: null>
                child 0, item: null
              cook-spinach_q2: list<item: string>
                child 0, item: string
              americano_q5: list<item: string>
                child 0, item: string
              cook-spinach_q3: list<item: string>
                child 0, item: string
              espresso_q6: list<item: null>
                child 0, item: null
              split_cookie_q2: list<item: string>
                child 0, item: string
              coffee_martini_q6: list<item: null>
                child 0, item: null
              keyboard_q1: list<item: string>
                child 0, item: string
              cut_roasted_beef_q1: list<item: string>
                child 0, item: string
              split_cookie_q5: list<item: null>
                child 0, item: null
              split_cookie_q8: list<item: null>
                child 0, item: null
              coffee_martini_q1: list<item: string>
                child 0, item: string
              americano_q2: list<item: string>
                child 0, item: string
              sear_steak_q1: list<item: string>
                child 0, item: string
              americano_q3: list<item: string>
                child 0, item: string
              cut_lemon_q2: list<item: string>
                child 0, item: string
              cut_roasted_beef_q7: list<item: string>
                child 0, item: string
              flame_salmon_q7: list<item: string>
                child 0, item: string
              cook-spinach_q6: list<item: null>
                child 0, item: null
              espresso_q4: list<item: string>
                child 0, item: string
              flame_steak_q3: list<item: string>
                child 0, item: string
              to
              {'cut_lemon_q1': List(Value('string')), 'cut_lemon_q2': List(Value('string')), 'cut_lemon_q3': List(Value('string')), 'cut_lemon_q4': List(Value('string')), 'cut_lemon_q5': List(Value('string')), 'cut_lemon_q6': List(Value('string')), 'cut_lemon_q7': List(Value('null')), 'cut_lemon_q8': List(Value('null')), 'espresso_q1': List(Value('string')), 'espresso_q2': List(Value('string')), 'espresso_q3': List(Value('string')), 'espresso_q4': List(Value('string')), 'espresso_q5': List(Value('null')), 'espresso_q6': List(Value('null')), 'espresso_q7': List(Value('string')), 'espresso_q8': List(Value('string')), 'keyboard_q1': List(Value('string')), 'keyboard_q2': List(Value('string')), 'keyboard_q3': List(Value('string')), 'keyboard_q4': List(Value('null')), 'keyboard_q5': List(Value('null')), 'keyboard_q6': List(Value('string')), 'torchchocolate_q1': List(Value('string')), 'torchchocolate_q2': List(Value('string')), 'torchchocolate_q3': List(Value('string')), 'torchchocolate_q4': List(Value('null')), 'torchchocolate_q5': List(Value('null')), 'torchchocolate_q6': List(Value('string')), 'torchchocolate_q7': List(Value('string')), 'cook-spinach_q1': List(Value('string')), 'cook-spinach_q2': List(Value('string')), 'cook-spinach_q3': List(Value('string')), 'cook-spinach_q4': List(Value('string')), 'cook-spinach_q5': List(Value('null')), 'cook-spinach_q6': List(Value('null')), 'cook-spinach_q7': List(Value('string')), 'cut_roasted_beef_q1': List(Value('string')), 'cut_roasted_beef_q2': List
              ...
              sear_steak_q3': List(Value('string')), 'sear_steak_q4': List(Value('string')), 'sear_steak_q5': List(Value('null')), 'sear_steak_q6': List(Value('null')), 'sear_steak_q7': List(Value('string')), 'split_cookie_q1': List(Value('string')), 'split_cookie_q2': List(Value('string')), 'split_cookie_q3': List(Value('string')), 'split_cookie_q4': List(Value('string')), 'split_cookie_q5': List(Value('null')), 'split_cookie_q6': List(Value('string')), 'split_cookie_q7': List(Value('string')), 'split_cookie_q8': List(Value('null')), 'americano_q1': List(Value('string')), 'americano_q2': List(Value('string')), 'americano_q3': List(Value('string')), 'americano_q4': List(Value('string')), 'americano_q5': List(Value('string')), 'americano_q6': List(Value('string')), 'americano_q7': List(Value('null')), 'americano_q8': List(Value('null')), 'americano_q9': List(Value('string')), 'americano_q10': List(Value('string')), 'coffee_martini_q1': List(Value('string')), 'coffee_martini_q2': List(Value('string')), 'coffee_martini_q3': List(Value('string')), 'coffee_martini_q4': List(Value('string')), 'coffee_martini_q5': List(Value('null')), 'coffee_martini_q6': List(Value('null')), 'coffee_martini_q7': List(Value('string')), 'flame_steak_q1': List(Value('string')), 'flame_steak_q2': List(Value('string')), 'flame_steak_q3': List(Value('string')), 'flame_steak_q4': List(Value('string')), 'flame_steak_q5': List(Value('null')), 'flame_steak_q6': List(Value('null')), 'flame_steak_q7': List(Value('string'))}
              because column names don't match

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.

R4D-Bench

Summary

R4D-Bench targets spatio-temporal referring segmentation in dynamic (4D) scenes: natural-language queries paired with pixel-accurate instance masks over time (COCO-style polygons or RLE, optional PNG unions), emphasizing motion, temporal relations (before / while / after), multi-target phrases, and distractor queries. The release uses three query tiers—36 main dense-GT queries, 89 extended (full per-scene set), and 246 supplementary English-only candidates without dense mask alignment. Mask IoU and track-ID evaluation live under evaluation/; full multi-view RGB, cameras, COLMAP, and Segment-then-Splat artifacts are not required for those shipped scripts. Scene imagery, where needed, should be obtained from Neu3D / HyperNeRF (or compatible) releases under their original licenses; the minimal annotation bundle does not include Segment-then-Splat pipeline outputs.

Statistics (at a glance)

Item	Location	Count / note
Dense mask GT — main release	`scripts/new_predictions_ground_truth_final.json`	36 queries (3 per scene × 12 scenes), per-frame `segmentation` and/or PNGs under `data/scenes/<scene>/query_masks/`
Dense mask GT — optional extension	e.g. `scripts/new_predictions_ground_truth_all_queries.json`	89 queries (sum of all entries in the 12 per-scene `*_queries.json` files)
Supplementary language candidates (no dense GT)	`data/queries/supplementary-queries/*.json`	246 English strings (`queries[].text` across files; auto-generated candidates; no dense `target_track_ids` alignment)
Evaluation metadata	`evaluation/R4D-Bench_queries.json`	89 entries — same query IDs as Tier B / `new_predictions_ground_truth_all_queries.json` (merged from per-scene `_queries.json`; order follows that dense GT file). `evaluation/R4D-Bench_predictions.json`* maps each `query_id` → `target_track_ids` for track-ID evaluation.

Query tiers (release semantics)

Tier	Description	Source	Dense mask GT?
A — Main	36 curated queries (3 / scene)	`scripts/predictions_ground_truth.py` without `--all-queries` → `scripts/new_predictions_ground_truth_final.json`	Yes
B — Extended	89 = every query in each scene’s `*_queries.json`	`scripts/predictions_ground_truth.py --all-queries` → e.g. `new_predictions_ground_truth_all_queries.json`	Yes
C — Supplementary	246 English phrases	`data/queries/supplementary-queries/`	No

Console shows “89 queries” only when running predictions_ground_truth.py with --all-queries, which sums all query_id entries across the 12 per-scene *_queries.json files (per-scene counts vary, e.g. americano 10, keyboard 6).

Reporting “200+” queries in a paper

If the paper claims “200+” by combining extended dense GT (89) with part of Tier C, a common breakdown is to count only the per-scene generated_queries_*.json files with 15 texts each (roughly 12 × 15 = 180), excluding longer Neu3D-prefixed files (e.g. 24 or 30 texts). That subset is not the same as the 246 total queries[].text strings in the repository; state exactly which files or tiers you count in the appendix or footnote.

Scenes (on-disk release)

There are 12 scene directories under data/scenes/:

americano, coffee_martini, cook_spinach, cut_lemon, cut_roasted_beef, espresso, flame_salmon, flame_steak, keyboard, sear_steak, split_cookie, torchchocolate

Naming conventions: Folder names use underscores. Some JSON fields and query_id prefixes use hyphens (e.g. cook-spinach, split-cookie). The scene cook_spinach ships files such as cook-spinach_queries.json and cook-spinach.json inside that folder.

Repository layout

R4D-Bench/
├── DATASET_LAYOUT.md               # Path checklist: core vs optional vs offline
├── README.md
├── evaluation/
│   ├── evaluate_mask_temporal.py   # mIoU, mAcc, temporal Acc, vIoU, …
│   ├── evaluate_r4dgs.py           # Track-id precision / recall / F1
│   ├── R4D-Bench_queries.json     # Unified query list (89) for eval + --queries-meta
│   └── R4D-Bench_predictions.json # query_id → target_track_ids (GT for track-id eval)
├── scripts/
│   ├── coco_scene_paths.py
│   ├── generate_instance_masks.py
│   ├── predictions_ground_truth.py
│   ├── enrich_ground_truth_with_mask_images.py
│   ├── new_predictions_ground_truth_final.json
│   ├── new_predictions_ground_truth_all_queries.json   # present if generated
│   └── *_with_paths.json                             # optional enriched variants
├── data/
│   ├── scenes/<scene>/          # images, COCO JSON, tracks, *_queries.json, query_masks/
│   ├── queries/
│   │   └── supplementary-queries/   # Tier C: extra English-only candidates (no dense GT)
│   ├── all_instance_masks/       # optional regeneratable snapshot
│   └── track_metadata.csv      # optional human-readable track_id → category (reference)
└── tools/                        # optional local utilities

If present, FINAL_REPORT.md / benchmark.md are maintainer notes and not required to use the benchmark.

Offline archive: Material that used to live under dataset_archive/ has been removed from this tree. Nothing here depends on that path. Benchmarks use data/scenes/, scripts/new_predictions_ground_truth_*.json, and evaluation/. A minimal distribution may omit data/all_instance_masks/ (regenerate with scripts/generate_instance_masks.py). See DATASET_LAYOUT.md for the full path map.

Query-related files (what to use when)

File	Role
`data/scenes/<scene>/*_queries.json`	Canonical source for each scene’s natural-language queries and `target_track_ids`. Used by `scripts/predictions_ground_truth.py` to build dense GT.
`scripts/new_predictions_ground_truth_final.json`	Dense mask GT for Tier A (36).
`scripts/new_predictions_ground_truth_all_queries.json`	Dense mask GT for Tier B (89) — full union of per-scene query lists.
`evaluation/R4D-Bench_queries.json`	Merged copy of all 89 queries (same IDs as Tier B, order aligned with `new_predictions_ground_truth_all_queries.json`). Pass to `--queries-meta` for per–`query_type` breakdown in mask metrics.
`evaluation/R4D-Bench_predictions.json`	Ground-truth `query_id` → `target_track_ids` for `evaluate_r4dgs.py` (replace with your model’s track-ID predictions when benchmarking).
`data/queries/supplementary-queries/*.json`	Tier C: auto-generated English phrases (`queries[].text`) only — 246 strings total; no dense masks or guaranteed track alignment.
`data/track_metadata.csv`	Optional spreadsheet mapping `track_id` → category and upstream dataset (HyperNeRF / Neu3D); not read by the shipped evaluation scripts—documentation / filtering only.

Legacy split lists under data/queries/ were removed to avoid stale duplicates; use the per-scene *_queries.json files and evaluation/R4D-Bench_queries.json instead.

Temporal and spatial alignment

Canonical time index is the integer frame_id in ground_truth.frames[] and existence_frames, and directory names frame_XXXXXX under data/scenes/<scene>/query_masks/.
Per-scene frame list comes from each scene’s COCO JSON (images[].file_name). Filenames may look like frame_000040.png or Roboflow-style names; scripts resolve resolution and ordering from COCO.
We do not enforce a single official Neu3D camera ID (e.g. cam00) or subsampling recipe for every scene. Users who pair this benchmark with original Neu3D / HyperNeRF downloads should align by visual / temporal correspondence; mask-only evaluation here depends only on the provided frame_id and mask geometry.

Mask-level evaluation

From the repository root:

# Sanity check: predictions = GT → metrics should be perfect
python evaluation/evaluate_mask_temporal.py --self-check \
  --queries-meta evaluation/R4D-Bench_queries.json

# Evaluate your prediction JSON (per-query, per-frame mask paths)
python evaluation/evaluate_mask_temporal.py \
  --predictions path/to/predictions.json \
  --output evaluation/mask_eval_report.json \
  --queries-meta evaluation/R4D-Bench_queries.json

Prediction JSON shape (examples): { "query_id": { "frames": { "1": "relative/or/abs/path.png", ... } } } or a list of { "frame_id", "mask_path" }. Paths resolve relative to --repo-root unless absolute.

Default --ground-truth is scripts/new_predictions_ground_truth_final.json (36 queries). To score all 89 extended queries, add
--ground-truth scripts/new_predictions_ground_truth_all_queries.json (predictions JSON must then cover the same query_id set you care about).

Useful options: --only-query-prefix <scene>, --only-queries, --debug-query <id>, --iou-threshold 0.5. See the docstring in evaluation/evaluate_mask_temporal.py.

Track-ID evaluation (set precision / recall / F1 on track IDs, no pixels):

python evaluation/evaluate_r4dgs.py \
  --queries evaluation/R4D-Bench_queries.json \
  --predictions evaluation/R4D-Bench_predictions.json \
  --output evaluation/track_id_eval_report.json

Replace --predictions with your model’s track-id output using the same query_id keys as in R4D-Bench_predictions.json.

Generating or refreshing masks and PNGs

scripts/generate_instance_masks.py reads each scene’s COCO JSON and writes one binary PNG per annotation instance under data/scenes/<scene>/instance_masks/<image_stem>/. That is per-instance rasterization—not the same as query-level union masks in new_predictions_ground_truth_final.json.

Required for a minimal release? No, if you ship queries + GT (segmentation and/or query_masks/) + evaluation code. Keep the script if you need to regenerate instance masks after editing COCO.

python scripts/generate_instance_masks.py --overwrite
python scripts/generate_instance_masks.py --scene americano --overwrite

python scripts/enrich_ground_truth_with_mask_images.py
python scripts/enrich_ground_truth_with_mask_images.py --only-query-prefix cut_lemon

Dependencies: Python 3.10+ recommended; numpy, Pillow. Optional: matplotlib, scikit-learn, wordcloud for auxiliary scripts.

Relationship to Segment-then-Splat (StS) and upstream data

Segment-then-Splat is a separate public pipeline (COLMAP, multi-view masks, Gaussian training, etc.). Typical StS directories (images/, sparse/, multiview_masks_*_merged/, PLY exports, …) are not part of the minimal R4D-Bench annotation release.
For StS-style training, follow their repository and obtain HyperNeRF / Neu3D (or compatible) imagery under the original licenses.
Roboflow provenance and URLs are under each scene’s README.dataset.txt / README.roboflow.txt (often CC BY 4.0 where stated—verify per scene).

Query schema (unified JSON)

Each entry in evaluation/R4D-Bench_queries.json includes among others:

query_id, scene, text_en, text_zh
target_track_ids, target_count
query_type (A / B / C), requires_motion_understanding
Optional: temporal_relation, expected_temporal_anchor, distractor, note

Items in scripts/new_predictions_ground_truth_final.json add ground_truth with target_tracks, existence_frames, and frames[].masks[] (segmentation, optional mask_image). Some question strings may be non-English depending on export version; canonical geometry is in segmentation / PNGs.

Citation

If you use R4D-Bench, cite this dataset repository and the original scene datasets (HyperNeRF, Neu3D, Roboflow sources as applicable). Add a BibTeX entry here after publication if desired.

License and third-party data

Annotations and code in this repository are released under the terms of the top-level LICENSE file when present; until then, treat usage as license-other and contact the maintainers if unsure.
Scene imagery and upstream assets remain under their original terms (Neu3D, HyperNeRF, Roboflow, etc.). Do not redistribute raw imagery unless the upstream license allows it.
Per-scene Roboflow metadata: data/scenes/<scene>/README.dataset.txt.

Contact

For questions about the benchmark definition, evaluation protocol, or file formats, open an issue in the project repository or contact the maintainers.

Downloads last month: 33

Total file size:

224 MB