SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2108.11743 · ICCV · 2021

Spatio-Temporal Dynamic Inference Network for Group Activity Recognition

Hangjie Yuan, Dong Ni, Mang Wang

arXiv · PDF · Open in the Atlas

Code that ran

4 of 12 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (1 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 4 have no contract verdict; 0 failures of our instrument; 8 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/2108.11743.json · badge: /api/ran/2108.11743.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
jacobyuan7/din_gar canonical (classifier label)
classifier basis: census
3 ran · 7 no recorded run
JacobYuan7/DIN-Group-Activity-Recognition-Benchmark community
Papers with Code archive
1 ran · 1 no recorded run
FunctionStatusWhere it lives
Dynamic_Person_Inference Ran
ran on a synthesized input · constructed an object; no result was computed
JacobYuan7/DIN-Group-Activity-Recognition-Benchmark/infer_module/dynamic_infer_module.py
code served (permissive licence) · get_code("0eed460a3795e49d")
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 an instance of its own class, Dynamic_Person_Inference: the run constructed the object and did not call it, so no result was computed. 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 MockConfig:
    def __init__(self):
        self.log_path = None

ARGS = {
    'in_dim': 4,
    'person_mat_shape': (3, 5),
    'stride': 1,
    'kernel_size': (3, 3),
    'dynamic_sampling': False,
    'sampling_ratio': [1],
    'group': 1,
    'scale_factor': False,
    'beta_factor': False,
    'parallel_inference': False,
    'cfg': MockConfig()
}
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 Dynamic_Person_Inference — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Dynamic_Person_Inference 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
collective_all_frames Ran
ran on a synthesized input
jacobyuan7/din_gar/collective.py
code served (permissive licence) · get_code("9042354c1505fdaa")
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 2, 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 torch
import numpy as np

torch.manual_seed(0)

# The function iterates over anns (treating it as a dict where keys are 's')
# and for each key 's', iterates over anns[s] (treating it as an iterable 'f')
# It returns a list of tuples (s, f)

# Create a simple dict structure where:
# - keys are strings (s)
# - values are iterables (lists, tuples, etc.) of frame identifiers (f)

anns = {
    "sequence_1": ["frame_0", "frame_1", "frame_2"],
    "sequence_2": ["frame_3", "frame_4"],
    "sequence_3": ["frame_5"]
}

ARGS = {
    "anns": anns
}
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 list, ndim 2, 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
collective_read_annotations Ran
ran on a synthesized input
jacobyuan7/din_gar/collective.py
code served (permissive licence) · get_code("4ace51a89549296c")
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 tempfile
import os

# Create a temporary directory and annotation file
temp_dir = tempfile.mkdtemp()
seq_dir = os.path.join(temp_dir, 'seq01')
os.makedirs(seq_dir, exist_ok=True)

# Create a sample annotations.txt file
# Format: frame_id x y w h action_id
annotations_path = os.path.join(seq_dir, 'annotations.txt')
with open(annotations_path, 'w') as f:
    # Frame 1 (frame_id % 10 == 1, so it will be included)
    f.write("1\t10\t20\t50\t60\t2\n")
    f.write("1\t100\t150\t60\t80\t3\n")
    # Frame 2 (not included in annotations)
    f.write("2\t15\t25\t55\t65\t2\n")
    # Frame 11 (frame_id % 10 == 1, so it will be included)
    f.write("11\t30\t40\t70\t90\t1\n")
    f.write("11\t120\t160\t65\t85\t2\n")

ARGS = {
    'path': temp_dir,
    'sid': 1
}
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
collective_read_dataset Ran
ran on a synthesized input
jacobyuan7/din_gar/collective.py
code served (permissive licence) · get_code("f237616dbb3d98b9")
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 torch
import numpy as np
import tempfile
import os
from collections import Counter

# Create a temporary directory structure with annotation files
temp_dir = tempfile.mkdtemp()

# Create annotation files for sequences
for sid in [1, 2, 3]:
    seq_dir = os.path.join(temp_dir, f'seq{sid:02d}')
    os.makedirs(seq_dir, exist_ok=True)
    
    # Create annotations.txt with sample data
    annotations_path = os.path.join(seq_dir, 'annotations.txt')
    with open(annotations_path, 'w') as f:
        # Write multiple frames with person detections
        # Format: frame_id x y w h action_id person_id
        for frame_num in range(1, 12):  # Frames 1-11
            for person_id in range(2):
                x, y, w, h = 10 + person_id * 50, 20, 40, 80
                action = 2  # Some action code
                f.write(f"{frame_num}\t{x}\t{y}\t{w}\t{h}\t{action}\t{person_id}\n")

# Define the arguments for collective_read_dataset
ARGS = {
    'path': temp_dir,
    'seqs': [1, 2, 3]
}
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 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
print_log Not yet run
unverified (harvested, no recorded run)
JacobYuan7/DIN-Group-Activity-Recognition-Benchmark/infer_module/dynamic_infer_module.py
code served (permissive licence) · get_code("6f3f9960fe07ddf5")
How it ran

no invocation recorded.

return_dataset Not yet run
unverified (harvested, no recorded run)
jacobyuan7/din_gar/dataset.py
code served (permissive licence) · get_code("50380e8c37e9f408")
How it ran

no invocation recorded.

test_volleyball Not yet run
unverified (harvested, no recorded run)
jacobyuan7/din_gar/train_net.py
code served (permissive licence) · get_code("2abf049f07d06c77")
How it ran

no invocation recorded.

test_volleyball Not yet run
unverified (harvested, no recorded run)
jacobyuan7/din_gar/train_net_dynamic.py
code served (permissive licence) · get_code("ef5bcfaacaa1e87e")
How it ran

no invocation recorded.

train_collective Not yet run
unverified (harvested, no recorded run)
jacobyuan7/din_gar/train_net.py
code served (permissive licence) · get_code("17d5b498d50e7ddd")
How it ran

no invocation recorded.

train_collective Not yet run
unverified (harvested, no recorded run)
jacobyuan7/din_gar/train_net_dynamic.py
code served (permissive licence) · get_code("1f8d5e5eca3b206f")
How it ran

no invocation recorded.

train_volleyball Not yet run
unverified (harvested, no recorded run)
jacobyuan7/din_gar/train_net.py
code served (permissive licence) · get_code("59684e160d4e51f8")
How it ran

no invocation recorded.

train_volleyball Not yet run
unverified (harvested, no recorded run)
jacobyuan7/din_gar/train_net_dynamic.py
code served (permissive licence) · get_code("04b1defe9f81fa3a")
How it ran

no invocation recorded.

Also present in other papers' code

Of 12 samples lifted from this paper's linked repositories, 12 carry a function identity and 0 do not. Of the 12 distinct functions with an identity, 2 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 2 later) and 10 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
train_collective 1 other paper
1 later
2602.03157 (2026-02)
Human-in-the-loop adaptation in group activity feature learning for team sports video retrieval
train_volleyball 1 other paper
1 later
2602.03157 (2026-02)
Human-in-the-loop adaptation in group activity feature learning for team sports video retrieval

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

Group activity recognition aims to understand the activity performed by a group of people. In order to solve it, modeling complex spatio-temporal interactions is the key. Previous methods are limited in reasoning on a predefined graph, which ignores the inherent person-specific interaction context. Moreover, they adopt inference schemes that are computationally expensive and easily result in the over-smoothing problem. In this paper, we manage to achieve spatio-temporal person-specific inferences by proposing Dynamic Inference Network (DIN), which composes of Dynamic Relation (DR) module and Dynamic Walk (DW) module. We firstly propose to initialize interaction fields on a primary spatio-temporal graph. Within each interaction field, we apply DR to predict the relation matrix and DW to predict the dynamic walk offsets in a jointprocessing manner, thus forming a person-specific interaction graph. By updating features on the specific graph, a person can possess a global-level interaction field with a local initialization. Experiments indicate both modules' effectiveness. Moreover, DIN 1 achieves significant improvement compared to previous state-of-the-art methods on two popular datasets under the same setting, while costing much less computation overhead of the reasoning module.

For agents

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

get_harvested_code_for_paper("2108.11743")
get_code_for_paper("2108.11743")
have("2108.11743")

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

curl https://syntology.ai/api/ran/2108.11743.json

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

[![Syntology run record](https://syntology.ai/api/ran/2108.11743.svg)](https://syntology.ai/paper/2108.11743)

Connect an agent — have() is free.