SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · aaai_28539 · AAAI · 2024

HR-Pro: Point-Supervised Temporal Action Localization via Hierarchical Reliability Propagation

Xiang Wang, Zhiwu Qing, Nong Sang, Xiaohao Xu, Huaxin Zhang, Changxin Gao

Open in the Atlas

Code that ran

9 of 12 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; 3 honoured the contract we drafted, 0 violated it and 5 have no contract verdict; 1 failure of our instrument; 3 no recorded run), measured 2026-11-22; 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/aaai_28539.json · badge: /api/ran/aaai_28539.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
pipixin321/HR-Pro canonical (classifier label)
classifier basis: census
9 ran · 3 no recorded run
FunctionStatusWhere it lives
color Ran
ran on a synthesized input
pipixin321/HR-Pro/log.py
code served (permissive licence) · get_code("e0abd0fd9d76a9db")
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 torch
import numpy as np
from termcolor import colored

torch.manual_seed(0)

ARGS = {
    'text': 'Hello World',
    'txt_color': 'green',
    'attrs': ['bold']
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, termcolor, torch — 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
compute_video_hit_at_k Ran
ran, honoured the contract we drafted
pipixin321/HR-Pro/eval/eval_classification.py
code served (permissive licence) · get_code("f914fa3c4de0cf80")
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 pandas as pd
import numpy as np

# Create ground truth dataframe
ground_truth_data = {
    'video-id': ['vid_1', 'vid_1', 'vid_2', 'vid_2', 'vid_3'],
    'label': ['cat', 'dog', 'bird', 'cat', 'dog']
}
ground_truth = pd.DataFrame(ground_truth_data)

# Create prediction dataframe
prediction_data = {
    'video-id': ['vid_1', 'vid_1', 'vid_1', 'vid_2', 'vid_2', 'vid_3'],
    'label': ['cat', 'dog', 'bird', 'cat', 'dog', 'cat'],
    'score': [0.9, 0.7, 0.5, 0.8, 0.6, 0.75]
}
prediction = pd.DataFrame(prediction_data)

ARGS = {
    'ground_truth': ground_truth,
    'prediction': prediction,
    'top_k': 3,
    'avg': False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas — 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
extract_region_feat Ran
ran on a synthesized input
pipixin321/HR-Pro/utils.py
code served (permissive licence) · get_code("08b1a1563a3fc401")
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. 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

torch.manual_seed(0)

# seq: binary sequence indicating regions [0,1,1,0,...,0,1,1,0]
# This should be a 1D tensor of 0s and 1s
seq = torch.tensor([0, 1, 1, 0, 1, 1, 1, 0], dtype=torch.long)

# embeded_feature: [T, F] where T is time steps and F is feature dimension
# T should match the length of seq
T = seq.shape[0]
F = 4
embeded_feature = torch.randn(T, F)

ARGS = {
    "seq": seq,
    "embeded_feature": embeded_feature,
}
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 list — a list 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
init_args Ran
ran on a synthesized input
pipixin321/HR-Pro/options.py
code served (permissive licence) · get_code("6e3cf888165cd63b")
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 Args. 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 os
import tempfile

class Args:
    def __init__(self):
        self.ckpt_path = tempfile.gettempdir()
        self.dataset = 'THUMOS14'
        self.task_info = 'test_task'
        self.act_thresh_cas = '[0.5, 0.6, 0.7]'
        self.act_thresh_agnostic = '[0.5]'
        self.lambdas = '[1.0, 1.0]'
        self.tIoU_thresh = '[0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7]'

ARGS = {
    'args': Args()
}
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 Args — a Args output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Args 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
interpolated_prec_rec Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
pipixin321/HR-Pro/eval/utils_eval.py
pointer only (licence: MIT); the stored copy get_code serves, from Ontheway361/BSN_pytorch, is pointer only (licence: NONE) · get_code("81aab88321de62c2")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from Ontheway361/BSN_pytorch, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

match_point Ran
ran on a synthesized input
pipixin321/HR-Pro/ranking.py
code served (permissive licence) · get_code("1ba203576f5260c5")
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

# Construct test data for match_point function

# props: dict mapping labels to lists of proposals
# Each proposal is a tuple: (start_time, end_time, confidence_score)
props = {
    "label_A": [
        (1.0, 5.0, 0.9),
        (2.0, 6.0, 0.7),
        (3.0, 7.0, 0.85),
    ],
    "label_B": [
        (0.5, 4.0, 0.8),
        (2.5, 5.5, 0.6),
    ],
}

# points: list of tuples
# Each point is a tuple: (time, label)
points = [
    (2.5, "label_A"),
    (3.5, "label_B"),
    (1.0, "label_A"),
]

ARGS = {
    "props": props,
    "points": points,
}
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 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
norm Ran
ran, honoured the contract we drafted
pipixin321/HR-Pro/utils.py
code served (permissive licence) · get_code("21d247685c39adc6")
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 Tensor, ndim 2, dtype family float, shape (2, 4) 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)

# The function 'norm' takes a single parameter 'data' which should be a tensor
# It computes L2 norm along the last dimension and normalizes by dividing
# So we need a tensor with at least 2 dimensions (batch, features)

data = torch.randn(2, 4)

ARGS = {
    "data": data
}
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 Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) 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
segment_iou Ran
ran, honoured the contract we drafted
pipixin321/HR-Pro/eval/utils_eval.py
pointer only (licence: MIT); the stored copy get_code serves, from Ontheway361/BSN_pytorch, is pointer only (licence: NONE) · get_code("d0744a2fe3151508")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from Ontheway361/BSN_pytorch, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

select_seed Ran
ran on a synthesized input
pipixin321/HR-Pro/utils.py
code served (permissive licence) · get_code("9699f4fd356595cf")
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 torch

torch.manual_seed(0)

# point_anno: shape (batch, time_steps, num_classes)
# - batch dimension: 2
# - time_steps dimension: 4
# - num_classes dimension: 3
# Used with: .max(dim=2)[0], .clone(), .detach(), shape indexing, item assignment
point_anno = torch.tensor([
    [[0.0, 0.5, 0.0], [0.0, 0.0, 0.0], [0.0, 0.0, 0.0], [0.8, 0.0, 0.0]],
    [[0.0, 0.7, 0.0], [0.0, 0.0, 0.0], [0.0, 0.0, 0.0], [0.6, 0.0, 0.0]]
])

# bkg_score: shape (batch, time_steps)
# - batch dimension: 2
# - time_steps dimension: 4
# Used with: indexing, comparison operations, argmax, shape indexing
bkg_score = torch.tensor([
    [0.1, 0.2, 0.3, 0.96],
    [0.05, 0.15, 0.25, 0.97]
])

ARGS = {
    "bkg_score": bkg_score,
    "point_anno": point_anno
}
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 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
compute_average_precision_classification Not yet run
unverified (harvested, no recorded run)
pipixin321/HR-Pro/eval/eval_classification.py
code served (permissive licence) · get_code("e1dfe4eceb3ff676")
How it ran

no invocation recorded.

compute_average_precision_detection Not yet run
unverified (harvested, no recorded run)
pipixin321/HR-Pro/eval/eval_detection.py
code served (permissive licence) · get_code("7be4ca6a451d8384")
How it ran

no invocation recorded.

get_blocked_videos Not yet run
unverified (harvested, no recorded run)
pipixin321/HR-Pro/eval/utils_eval.py
code served (permissive licence) · get_code("a01a64a508273392")
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, 8 are also present in other papers whose code we hold (8 with at least one paper dated earlier, 0 the same month, 1 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.

FunctionAlso present inEarliest-datedOthers named
interpolated_prec_rec 13 other papers
10 dated earlier, 3 same year, order not determinable
1804.05113 (2018-04)
Multilevel Language and Vision Integration for Text-to-Clip Retrieval
1806.02964 (2018-06)
BSN: Boundary Sensitive Network for Temporal Action Proposal Generation
1907.09702 (2019-07)
BMN: Boundary-Matching Network for Temporal Action Proposal Generation
and 10 more not named here
segment_iou 13 other papers
10 dated earlier, 3 same year, order not determinable
1804.05113 (2018-04)
Multilevel Language and Vision Integration for Text-to-Clip Retrieval
1806.02964 (2018-06)
BSN: Boundary Sensitive Network for Temporal Action Proposal Generation
1907.09702 (2019-07)
BMN: Boundary-Matching Network for Temporal Action Proposal Generation
and 10 more not named here
compute_average_precision_detection 4 other papers
4 dated earlier
1911.11462 (2019-11)
G-TAD: Sub-Graph Localization for Temporal Action Detection
aaai_6793 (2020 (year only))
Background Suppression Network for Weakly-Supervised Temporal Action Localization
aaai_16280 (2021 (year only))
Weakly-supervised Temporal Action Localization by Uncertainty Modeling
and 1 more not named here
get_blocked_videos 4 other papers
4 dated earlier
1911.11462 (2019-11)
G-TAD: Sub-Graph Localization for Temporal Action Detection
aaai_6793 (2020 (year only))
Background Suppression Network for Weakly-Supervised Temporal Action Localization
aaai_16280 (2021 (year only))
Weakly-supervised Temporal Action Localization by Uncertainty Modeling
and 1 more not named here
norm 3 other papers
1 dated earlier, 1 later, 1 same year, order not determinable
2302.05160 (2023-02)
Dual Memory Units with Uncertainty Regulation for Weakly Supervised Video Anomaly Detection
2412.06171 (2024-12)
Holmes-VAU: Towards Long-term Video Anomaly Understanding at Any Granularity
compute_video_hit_at_k 2 other papers
2 dated earlier
aaai_16280 (2021 (year only))
Weakly-supervised Temporal Action Localization by Uncertainty Modeling
aaai_19900 (2022 (year only))
DCAN: Improving Temporal Action Detection via Dual Context Aggregation
color 1 other paper
1 dated earlier
136730727 (2022 (year only))
Self-Supervised Sparse Representation for Video Anomaly Detection
compute_average_precision_classification 1 other paper
1 dated earlier
aaai_16280 (2021 (year only))
Weakly-supervised Temporal Action Localization by Uncertainty Modeling

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

Point-supervised Temporal Action Localization (PSTAL) is an emerging research direction for label-efficient learning. However, current methods mainly focus on optimizing the network either at the snippet-level or the instance-level, neglecting the inherent reliability of point annotations at both levels. In this paper, we propose a Hierarchical Reliability Propagation (HR-Pro) framework, which consists of two reliability-aware stages: Snippet-level Discrimination Learning and Instance-level Completeness Learning, both stages explore the efficient propagation of high-confidence cues in point annotations. For snippet-level learning, we introduce an online-updated memory to store reliable snippet prototypes for each class. We then employ a Reliability-aware Attention Block to capture both intra-video and inter-video dependencies of snippets, resulting in more discriminative and robust snippet representation. For instance-level learning, we propose a point-based proposal generation approach as a means of connecting snippets and instances, which produces high-confidence proposals for further optimization at the instance level. Through multi-level reliability-aware learning, we obtain more reliable confidence scores and more accurate temporal boundaries of predicted proposals. Our HR-Pro achieves state-of-the-art performance on multiple challenging benchmarks, including an impressive average mAP of 60.3% on THUMOS14. Notably, our HR-Pro largely surpasses all previous point-supervised methods, and even outperforms several competitive fully-supervised methods. Code will be available at https://github.com/pipixin321/HR-Pro.

For agents

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

get_harvested_code_for_paper("aaai_28539")
get_code_for_paper("aaai_28539")
have("aaai_28539")

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

curl https://syntology.ai/api/ran/aaai_28539.json

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

[![Syntology run record](https://syntology.ai/api/ran/aaai_28539.svg)](https://syntology.ai/paper/aaai_28539)

Connect an agent — have() is free.