SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2203.01577 · CVPR · 2022

HOI4D: A 4D Egocentric Dataset for Category-Level Human-Object Interaction

Hao Shen, Li Yi, He Wang, Weikang Wan, Che Jiang, Yunze Liu, Kangbo Lyu, Zhoujie Fu, Boqiang Liang

arXiv · PDF · Open in the Atlas

Code that ran

5 of 5 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 5 have no contract verdict; 0 failures of our instrument; 0 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2203.01577.json · badge: /api/ran/2203.01577.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
leolyliu/HOI4D-Instructions canonical (classifier label)
classifier basis not recorded on the link
5 ran · 0 no recorded run
FunctionStatusWhere it lives
get_color_map Ran
ran on a synthesized input
leolyliu/HOI4D-Instructions/prepare_4Dseg/palette.py
code served (permissive licence) · get_code("bfda9026ff9b0fa9")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 2, dtype family int, shape (256, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

ARGS = {
    'N': 256
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family int, shape (256, 3) — shape (256, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_color_map Ran
ran on a synthesized input
leolyliu/HOI4D-Instructions/utils/color_map.py
code served (permissive licence) · get_code("208ae9d669dc0863")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 2, dtype family int, shape (256, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

ARGS = {
    "N": 256
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family int, shape (256, 3) — shape (256, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_csv Ran
ran on a synthesized input
leolyliu/HOI4D-Instructions/prepare_4Dseg/utils.py
code served (permissive licence) · get_code("cf2e73337e63c1c8")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned reader. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import csv
import tempfile
import os

# Create a temporary CSV file for testing
temp_dir = tempfile.mkdtemp()
csv_path = os.path.join(temp_dir, "test.csv")

# Write some sample data to the CSV file
with open(csv_path, 'w', newline='') as f:
    writer = csv.writer(f)
    writer.writerow(['col1', 'col2', 'col3'])
    writer.writerow(['a', 'b', 'c'])
    writer.writerow(['d', 'e', 'f'])

ARGS = {
    'path': csv_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type reader — a reader output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a reader output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_foreground_depths_and_labels Ran
ran on a synthesized input
leolyliu/HOI4D-Instructions/prepare_4Dseg/scene2frame.py
code served (permissive licence) · get_code("1be362a930f4503e")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import cv2
import numpy as np

# Create temporary depth and mask images in memory
# The function uses cv2.imread, so we need to create actual image files
# But since we can't do file I/O, we'll create minimal mock objects

class MockImage:
    def __init__(self, shape, dtype=np.uint16):
        self.data = np.zeros(shape, dtype=dtype)
    
    def copy(self):
        return self.data.copy()
    
    def __getitem__(self, key):
        return self.data[key]
    
    def __setitem__(self, key, value):
        self.data[key] = value
    
    @property
    def shape(self):
        return self.data.shape

# Create depth image (1080, 1920) as uint16
depth_array = np.zeros((1080, 1920), dtype=np.uint16)

# Create mask image (1080, 1920, 3) as uint8
mask_array = np.zeros((1080, 1920, 3), dtype=np.uint8)

# Create boolean mask arrays for ds parameter
# These are used with ~d syntax (bitwise NOT), so they should be boolean arrays
d1 = np.random.rand(1080, 1920) > 0.5  # Random boolean mask
d2 = np.random.rand(1080, 1920) > 0.5  # Another random boolean mask
ds = [d1, d2]

# Create simple list labels
ls = [0, 1, 2]

# Create instance segmentation labels
ls_instanceseg = [10, 20, 30]

# Mock cv2.imread to return our arrays
original_imread = cv2.imread

def mock_imread(path, flags):
    if "depth" in path:
        return depth_array.copy()
    elif "mask" in path:
        return mask_array.copy()
    return None

cv2.imread = mock_imread

ARGS = {
    "depth_path": "dummy_depth.png",
    "mask_path": "dummy_mask.png",
    "ds": ds,
    "ls": ls,
    "ls_instanceseg": ls_instanceseg
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
prepare_datalists Ran
ran on a synthesized input
leolyliu/HOI4D-Instructions/prepare_4Dseg/prepare_datalists.py
code served (permissive licence) · get_code("90f164a947bf582d")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import os
import tempfile
import shutil

# Create a temporary directory structure matching the nested hierarchy
# that prepare_datalists expects
temp_root = tempfile.mkdtemp()

# Create the nested directory structure: ZY/H/C/N/S/s/T/3Dseg/
zy_dir = os.path.join(temp_root, "ZY_sample")
os.makedirs(zy_dir)

h_dir = os.path.join(zy_dir, "H_sample")
os.makedirs(h_dir)

c_dir = os.path.join(h_dir, "C_sample")
os.makedirs(c_dir)

n_dir = os.path.join(c_dir, "N_sample")
os.makedirs(n_dir)

s_capital_dir = os.path.join(n_dir, "S_sample")
os.makedirs(s_capital_dir)

s_lower_dir = os.path.join(s_capital_dir, "s_sample")
os.makedirs(s_lower_dir)

t_dir = os.path.join(s_lower_dir, "T_sample")
os.makedirs(t_dir)

# Create the required 3Dseg directory with label.txt file
seg_dir = os.path.join(t_dir, "3Dseg")
os.makedirs(seg_dir)

label_file = os.path.join(seg_dir, "label.txt")
with open(label_file, "w") as f:
    f.write("label data")

ARGS = {
    "root_dir": temp_root
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 4 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 4 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

We present HOI4D, a large-scale 4D egocentric dataset with rich annotations, to catalyze the research of categorylevel human-object interaction. HOI4D consists of 2.4M RGB-D egocentric video frames over 4000 sequences collected by 9 participants interacting with 800 different object instances from 16 categories over 610 different indoor rooms. Frame-wise annotations for panoptic segmentation, motion segmentation, 3D hand pose, category-level object pose and hand action have also been provided, together with reconstructed object meshes and scene point clouds. With HOI4D, we establish three benchmarking tasks to promote category-level HOI from 4D visual signals including semantic segmentation of 4D dynamic point cloud sequences, category-level object pose tracking, and egocentric action segmentation with diverse interaction targets. In-depth analysis shows HOI4D poses great challenges to existing methods and produces huge research opportunities.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2203.01577")
get_code_for_paper("2203.01577")
have("2203.01577")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2203.01577.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2203.01577.svg)](https://syntology.ai/paper/2203.01577)

Connect an agent — have() is free.