id stringlengths 7 22 | video unknown | instruction stringlengths 0 148 | segments listlengths 2 58 | metadata stringlengths 435 735 |
|---|---|---|---|---|
homer_1 | "AAAAIGZ0eXBpc29tAAACAGlzb21pc28yYXZjMW1wNDEAAAAIZnJlZQF/OEFtZGF0AAAAHgYFGkdWStxcTEM/lO/FETzRQ6gB/8z(...TRUNCATED) | use pitcher to pour water into tall glass and wine glass | [{"start_sec":0.0,"end_sec":7.709,"subtask":"twist open the pitcher lid"},{"start_sec":7.709,"end_se(...TRUNCATED) | "{\"local_reannotation_line\": 62, \"local_reannotation_saved_at_unix\": 1782483511.842509, \"num_se(...TRUNCATED) |
homer_2 | "AAAAIGZ0eXBpc29tAAACAGlzb21pc28yYXZjMW1wNDEAAAAIZnJlZQKrCVBtZGF0AAB0nGWIgABAAAJ/myGzVBf4QHEs7t40P2J(...TRUNCATED) | wash the dishes in the sink | [{"start_sec":0.0,"end_sec":3.981,"subtask":"turn on the kitchen faucet"},{"start_sec":3.981,"end_se(...TRUNCATED) | "{\"local_reannotation_line\": 63, \"local_reannotation_saved_at_unix\": 1782483513.340191, \"num_se(...TRUNCATED) |
homer_3 | "AAAAIGZ0eXBpc29tAAACAGlzb21pc28yYXZjMW1wNDEAAAAIZnJlZQGKw5BtZGF0AABREWW4Qd/jlrk4INj/+Ej8OiMPBCMGRbc(...TRUNCATED) | use a spoon to stir contents in a bowl | [{"start_sec":0.0,"end_sec":3.285,"subtask":"pick up the green bag with brown powder"},{"start_sec":(...TRUNCATED) | "{\"local_reannotation_line\": 64, \"local_reannotation_saved_at_unix\": 1782483515.3551092, \"num_s(...TRUNCATED) |
homer_4 | "AAAAIGZ0eXBpc29tAAACAGlzb21pc28yYXZjMW1wNDEAAAAIZnJlZQZHhwVtZGF0AAAC2QYF///V3EXpvebZSLeWLNgg2SPu73g(...TRUNCATED) | Clean tables, desks, or shelves with a cloth | [{"start_sec":0.0,"end_sec":9.69,"subtask":"wipe the top surface of the wooden nightstand with a whi(...TRUNCATED) | "{\"local_reannotation_line\": 66, \"local_reannotation_saved_at_unix\": 1782483781.602892, \"num_se(...TRUNCATED) |
homer_5 | "AAAAIGZ0eXBpc29tAAACAGlzb21pc28yYXZjMW1wNDEAAAAIZnJlZQEe2YhtZGF0AAAAOgYFMkdWStxcTEM/lO/FETzRQ6gBAAA(...TRUNCATED) | use a cloth to dust white shelf | [{"start_sec":0.0,"end_sec":4.659,"subtask":"wipe the the top of white shelf with a purple cloth"},{(...TRUNCATED) | "{\"local_reannotation_line\": 59, \"local_reannotation_saved_at_unix\": 1782483133.715023, \"num_se(...TRUNCATED) |
homer_7 | "AAAAIGZ0eXBpc29tAAACAGlzb21pc28yYXZjMW1wNDEAAAAIZnJlZQE+j19tZGF0AAAAOgYFMkdWStxcTEM/lO/FETzRQ6gBAAA(...TRUNCATED) | draw simple beach and ocean scene using a pencil | [{"start_sec":0.0,"end_sec":1.333,"subtask":"pick up the pencil"},{"start_sec":1.333,"end_sec":21.55(...TRUNCATED) | "{\"local_reannotation_line\": 68, \"local_reannotation_saved_at_unix\": 1782483887.59743, \"num_seg(...TRUNCATED) |
homer_9 | "AAAAIGZ0eXBpc29tAAACAGlzb21pc28yYXZjMW1wNDEAAAAIZnJlZQGPOb1tZGF0AAAAHgYFGkdWStxcTEM/lO/FETzRQ6gB/8z(...TRUNCATED) | wash the two glasses in the sink | [{"start_sec":2.817,"end_sec":5.473,"subtask":"turn on the kitchen faucet"},{"start_sec":5.473,"end_(...TRUNCATED) | "{\"local_reannotation_line\": 70, \"local_reannotation_saved_at_unix\": 1782484600.948663, \"num_se(...TRUNCATED) |
homer_10 | "AAAAIGZ0eXBpc29tAAACAGlzb21pc28yYXZjMW1wNDEAAAAIZnJlZQEaGdhtZGF0AAAAHgYFGkdWStxcTEM/lO/FETzRQ6gB/8z(...TRUNCATED) | clean the countertop and stovetop with a cloth | [{"start_sec":0.0,"end_sec":1.334,"subtask":"unfold the yellow cloth"},{"start_sec":1.334,"end_sec":(...TRUNCATED) | "{\"local_reannotation_line\": 71, \"local_reannotation_saved_at_unix\": 1782485062.921957, \"num_se(...TRUNCATED) |
homer_11 | "AAAAIGZ0eXBpc29tAAACAGlzb21pc28yYXZjMW1wNDEAAAAIZnJlZQFuEoBtZGF0AABrVmW4QKchAx6BHQaFg8WX5ocRymIbOdh(...TRUNCATED) | open, inspect and close all containers, bottles and jars in front of you | [{"start_sec":0.0,"end_sec":3.421,"subtask":"pick up small jar containing coffee beans from the bott(...TRUNCATED) | "{\"local_reannotation_line\": 73, \"local_reannotation_saved_at_unix\": 1782486078.376809, \"num_se(...TRUNCATED) |
homer_12 | "AAAAIGZ0eXBpc29tAAACAGlzb21pc28yYXZjMW1wNDEAAAAIZnJlZQEpNDFtZGF0AABAE2W4QJe0mU6BnD//CN8SdZcYGuUFdER(...TRUNCATED) | open, inspect and close all containers, bottles and jars in front of you | [{"start_sec":0.0,"end_sec":2.617,"subtask":"pick up the blue plastic container from the counter and(...TRUNCATED) | "{\"local_reannotation_line\": 75, \"local_reannotation_saved_at_unix\": 1782486638.407902, \"num_se(...TRUNCATED) |
WGO-Bench: What's Going On Benchmark
WGO-Bench is a small, manually annotated benchmark for evaluating how well vision-language models can turn robot and egocentric manipulation videos into timestamped subtask annotations.
Each row contains one video episode, a high-level task instruction, and gold subtask segments with start time, end time, and a concise action label. The benchmark is designed for two related tasks:
- Boundary detection: predict where one meaningful manipulation subtask ends and the next begins.
- Subtask labeling: given a fixed segment, describe the completed manipulation event.
Dataset Summary
| Metric | Value |
|---|---|
| Episodes | 100 |
| Gold subtask segments | 743 |
| Unique task instructions | 63 |
| Source families | 3 |
| Format | Parquet |
| Video storage | MP4 bytes embedded in each row |
| Robot-state coverage | 75 robotic episodes; HomER is video-only |
Sources
| Source | Episodes |
|---|---|
| HomER egocentric videos | 25 |
| RoboInter DROID station-arm videos | 50 |
| RoboCOIN Galaxea R1 Lite head-camera videos | 25 |
Schema
| Column | Type | Description |
|---|---|---|
id |
string | Stable episode identifier. |
video |
binary | MP4 video bytes embedded directly in the parquet row. |
instruction |
string | High-level task instruction for the episode. |
segments |
list | Gold {start_sec, end_sec, subtask} annotations. |
metadata |
string | JSON metadata with source-specific fields. |
perception_state |
struct or null | Synchronized, named robot-state channels for DROID/Galaxea; null for HomER. |
DROID retains EEF position, rotation, and gripper state. Galaxea retains
bilateral arm joints and grippers plus its available EEF- and gripper-derived
channels. Actions are not included. HomER has no robot telemetry, so its rows
contain null rather than synthetic zeros.
Quick Start
from datasets import load_dataset
dataset = load_dataset("macrodata/WGO-Bench", split="train")
example = dataset[0]
print(example["instruction"])
print(example["segments"])
print(example["perception_state"])
To use the parquet directly:
import pandas as pd
df = pd.read_parquet("hf://datasets/macrodata/WGO-Bench/data/annotations.parquet")
Rebuilding Robot States
scripts/perception_sources.lock.json pins both annotation maps and all 26
upstream robot-dataset revisions. Rebuild the enriched table from the original
annotations Parquet:
python scripts/embed_perception_states.py \
data/annotations.original.parquet \
data/annotations.parquet
Requirements are Python 3.11+, huggingface_hub, and pyarrow. The exporter
validates episode coverage, frame counts, channel widths, and timestamp order,
then writes a SHA-256 provenance file next to the enriched Parquet. Hugging Face
caches all state sources; pass --cache-dir to use a shared cache. The DROID
state data comes from one roughly 241 MB upstream tar archive. The embedded
state payload adds roughly 1.5 MB before Parquet-level recompression.
Annotation Policy
Segments are intended to describe completed manipulation events, not every small pose adjustment. A new segment should generally correspond to a visible state change such as picking up, placing, opening, closing, moving, pouring, wiping, cutting, or transferring an object.
The labels are short action descriptions grounded in what is visible in the clip. They are meant to be useful for downstream policy training, reward modeling, and evaluating automatic subtask annotation pipelines.
Related Resources
- Blog post: Segmenting Robot Video into Actionable Subtasks
- Refiner example: Subtask annotations
- Code: macrodata-labs/refiner
- Downloads last month
- 1,387