SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2007.08728 · 2020

Detecting Human-Object Interactions with Action Co-occurrence Priors

arXiv · PDF · Open in the Atlas

Code that ran

10 of 10 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 10 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/2007.08728.json · badge: /api/ran/2007.08728.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
Dong-JinKim/ActionCooccurrencePriors canonical (classifier label)
classifier basis: census
10 ran · 0 no recorded run
FunctionStatusWhere it lives
compute_area Ran
ran on a synthesized input
Dong-JinKim/ActionCooccurrencePriors/utils/bbox_utils.py
code served (permissive licence) · get_code("5f8551dd0fb2cf55")
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 int, ndim 0, dtype family int, shape () 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 torch

torch.manual_seed(0)

# bbox is a tuple of 4 coordinates (x1, y1, x2, y2)
bbox = (10, 20, 30, 40)

# invalid is the value to return if the bbox is invalid (x2 <= x1 or y2 <= y1)
invalid = None

ARGS = {
    "bbox": bbox,
    "invalid": invalid,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int, ndim 0, dtype family int, shape () — shape () 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
compute_area_batch Ran
ran on a synthesized input
Dong-JinKim/ActionCooccurrencePriors/utils/bbox_utils.py
code served (permissive licence) · get_code("5c81f0f6cf474579")
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 1, dtype family float, shape (2,) 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

# Create a small batch of bounding boxes
# Format: [x1, y1, x2, y2] for each box
# batch_size = 2, each box has 4 coordinates
bbox = np.array([
    [10.0, 20.0, 50.0, 80.0],  # valid box
    [100.0, 150.0, 90.0, 140.0]  # invalid box (x2 < x1, y2 < y1)
], dtype=np.float32)

ARGS = {
    "bbox": bbox
}
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 1, dtype family float, shape (2,) — shape (2,) 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
compute_iou Ran
ran on a synthesized input
Dong-JinKim/ActionCooccurrencePriors/utils/bbox_utils.py
code served (permissive licence) · get_code("d342aa66e09cb8a9")
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 float, ndim 0, dtype family float, shape () 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 torch
import numpy as np

torch.manual_seed(0)

# The function compute_iou takes two bounding boxes and an optional verbose flag
# bbox1 and bbox2 are expected to be unpacked as x1, y1, x2, y2
# They should be iterables/tuples with 4 elements

bbox1 = [10, 20, 50, 60]  # x1, y1, x2, y2 for first box
bbox2 = [30, 40, 70, 80]  # x1, y1, x2, y2 for second box

ARGS = {
    "bbox1": bbox1,
    "bbox2": bbox2,
    "verbose": False,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float, ndim 0, dtype family float, shape () — shape () 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
create_mlp Ran
ran on a synthesized input
Dong-JinKim/ActionCooccurrencePriors/utils/pytorch_layers.py
code served (permissive licence) · get_code("00ad97d3fa0e9432")
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 MLP. 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 torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    'const': {
        'in_dim': 4,
        'layer_units': [8, 6],
        'activation': 'ReLU',
        'drop_prob': 0.1
    }
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type MLP — a MLP output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a MLP 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
dumps_json_object Ran
ran on a synthesized input
Dong-JinKim/ActionCooccurrencePriors/utils/io.py
code served (permissive licence) · get_code("6c82cba89895242a")
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 str. 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 json
import numpy as np

class NumpyAwareJSONEncoder(json.JSONEncoder):

    def default(self, obj):
        if isinstance(obj, np.ndarray):
            if obj.ndim == 1:
                return obj.tolist()
            else:
                return [self.default(obj[i]) for i in range(obj.shape[0])]
        elif isinstance(obj, np.int64):
            return int(obj)
        elif isinstance(obj, np.int32):
            return int(obj)
        elif isinstance(obj, np.int16):
            return int(obj)
        elif isinstance(obj, np.float64):
            return float(obj)
        elif isinstance(obj, np.float32):
            return float(obj)
        elif isinstance(obj, np.float16):
            return float(obj)
        elif isinstance(obj, np.uint64):
            return int(obj)
        elif isinstance(obj, np.uint32):
            return int(obj)
        elif isinstance(obj, np.uint16):
            return int(obj)
        return json.JSONEncoder.default(self, obj)

def dumps_json_object(dump_object, indent=4):
    data = json.dumps(dump_object, cls=NumpyAwareJSONEncoder, sort_keys=True, indent=indent)
    return data

# Create test data with various numpy types
dump_object = {
    "int64_val": np.int64(42),
    "int32_val": np.int32(10),
    "int16_val": np.int16(5),
    "float64_val": np.float64(3.14),
    "float32_val": np.float32(2.71),
    "float16_val": np.float16(1.5),
    "uint64_val": np.uint64(100),
    "uint32_val": np.uint32(50),
    "uint16_val": np.uint16(25),
    "array_1d": np.array([1, 2, 3, 4]),
    "array_2d": np.array([[1, 2], [3, 4], [5, 6]]),
    "nested": {
        "inner_array": np.array([7, 8, 9]),
        "inner_value": np.float32(1.23)
    },
    "list": [1, 2, "string", np.int64(99)],
    "string": "test",
    "number": 42
}

indent_value = 2

ARGS = {
    "dump_object": dump_object,
    "indent": indent_value
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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_activation Ran
ran on a synthesized input
Dong-JinKim/ActionCooccurrencePriors/utils/pytorch_layers.py
code served (permissive licence) · get_code("0c0c242e0bf98442")
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 ReLU. 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 torch
import torch.nn as nn

torch.manual_seed(0)

class Identity(nn.Module):
    def __init__(self):
        super(Identity, self).__init__()

    def forward(self, x):
        return x

def get_activation(name):
    if name == 'ReLU':
        return nn.ReLU(inplace=True)
    elif name == 'Tanh':
        return nn.Tanh()
    elif name == 'Identity':
        return Identity()
    elif name == 'Sigmoid':
        return nn.Sigmoid()
    elif name == 'LeakyReLU':
        return nn.LeakyReLU(0.2, inplace=True)
    else:
        assert False, 'Not Implemented'

ARGS = {
    'name': 'ReLU'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ReLU — a ReLU output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ReLU 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
load_json_object Ran
ran on a synthesized input
Dong-JinKim/ActionCooccurrencePriors/utils/io.py
code served (permissive licence) · get_code("fbb293b5e43e3fa4")
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 dict. 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 json
import gzip
import tempfile
import os

# Create a temporary directory for test files
temp_dir = tempfile.mkdtemp()

# Create a simple JSON file (uncompressed)
uncompressed_file = os.path.join(temp_dir, "test.json")
test_data = {"key": "value", "number": 42, "list": [1, 2, 3]}
with open(uncompressed_file, 'w') as f:
    json.dump(test_data, f)

# Create a compressed JSON file
compressed_file = os.path.join(temp_dir, "test.json.gz")
with gzip.open(compressed_file, 'wb') as f:
    f.write(json.dumps(test_data).encode('utf8'))

# Define ARGS for the function
ARGS = {
    'file_name': uncompressed_file,
    'compress': False
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
load_pickle_object Ran
ran on a synthesized input
Dong-JinKim/ActionCooccurrencePriors/utils/io.py
code served (permissive licence) · get_code("f9531bd8a2f31c23")
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 dict. 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 pickle
import gzip
import tempfile
import os

# Create a temporary directory for test files
temp_dir = tempfile.mkdtemp()

# Create test data to pickle
test_data = {
    'key1': 'value1',
    'key2': [1, 2, 3],
    'key3': {'nested': 'dict'}
}

# Create a compressed pickle file
compressed_file = os.path.join(temp_dir, 'test_compressed.pkl.gz')
with gzip.open(compressed_file, 'wb') as f:
    f.write(pickle.dumps(test_data))

# Create an uncompressed pickle file
uncompressed_file = os.path.join(temp_dir, 'test_uncompressed.pkl')
with open(uncompressed_file, 'wb') as f:
    f.write(pickle.dumps(test_data))

# Define arguments for the function
ARGS = {
    'file_name': compressed_file,
    'compress': True
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
manage_required_args Ran
ran on a synthesized input
Dong-JinKim/ActionCooccurrencePriors/utils/argparse_utils.py
code served (permissive licence) · get_code("fd54712b1bf5feeb")
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 argparse
import sys

# Create a minimal args object
class Args:
    def __init__(self):
        self.arg1 = "value1"
        self.arg2 = None
        self.arg3 = "value3"

args = Args()

# Create a parser with some options
parser = argparse.ArgumentParser()
parser.add_argument('--arg1', help='First argument', default=None)
parser.add_argument('--arg2', help='Second argument (required)', default=None)
parser.add_argument('--arg3', help='Third argument with choices', choices=['value3', 'value4'], default=None)

ARGS = {
    'args': args,
    'parser': parser,
    'required_args': ['arg2'],
    'optional_args': ['arg1', 'arg3'],
    'exit_if_unspecified': False
}
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
str_to_bool Ran
ran on a synthesized input
Dong-JinKim/ActionCooccurrencePriors/utils/argparse_utils.py
code served (permissive licence) · get_code("8d87f08966a2f30c")
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 bool, ndim 0, dtype family float, shape () 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 argparse

def str_to_bool(v):
    if v.lower() in ('yes', 'true', 't', 'y', '1'):
        return True
    elif v.lower() in ('no', 'false', 'f', 'n', '0'):
        return False
    else:
        raise argparse.ArgumentTypeError('Boolean value expected.')

ARGS = {
    'v': 'true'
}
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 bool, ndim 0, dtype family float, shape () — shape () 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

Also present in other papers' code

Of 10 samples lifted from this paper's linked repositories, 10 carry a function identity and 0 do not. Of the 10 distinct functions with an identity, 7 are also present in other papers whose code we hold (7 with at least one paper dated earlier, 0 the same month, 1 later) and 3 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.

FunctionAlso present inEarliest-datedOthers named
compute_area 2 other papers
1 dated earlier, 1 later
2003.05541 (2020-03)
VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions
2112.08647 (2021-12)
QAHOI: Query-Based Anchors for Human-Object Interaction Detection
compute_area_batch 1 other paper
1 dated earlier
2003.05541 (2020-03)
VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions
compute_iou 1 other paper
1 dated earlier
2003.05541 (2020-03)
VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions
dumps_json_object 1 other paper
1 dated earlier
2003.05541 (2020-03)
VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions
load_json_object 1 other paper
1 dated earlier
2003.05541 (2020-03)
VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions
load_pickle_object 1 other paper
1 dated earlier
2003.05541 (2020-03)
VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions
str_to_bool 1 other paper
1 dated earlier
1703.05593 (2017-03)
Convolutional neural network architecture for geometric matching

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

A common problem in human-object interaction (HOI) detection task is that numerous HOI classes have only a small number of labeled examples, resulting in training sets with a long-tailed distribution. The lack of positive labels can lead to low classification accuracy for these classes. Towards addressing this issue, we observe that there exist natural correlations and anti-correlations among human-object interactions. In this paper, we model the correlations as action co-occurrence matrices and present techniques to learn these priors and leverage them for more effective training, especially in rare classes. The utility of our approach is demonstrated experimentally, where the performance of our approach exceeds the state-of-the-art methods on both of the two leading HOI detection benchmark datasets, HICO-Det and V-COCO.

For agents

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

get_harvested_code_for_paper("2007.08728")
get_code_for_paper("2007.08728")
have("2007.08728")

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

curl https://syntology.ai/api/ran/2007.08728.json

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

[![Syntology run record](https://syntology.ai/api/ran/2007.08728.svg)](https://syntology.ai/paper/2007.08728)

Connect an agent — have() is free.