SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2009.07641 · 2020

BSN++: Complementary Boundary Regressor with Scale-Balanced Relation Modeling for Temporal Action Proposal Generation

arXiv · PDF · Open in the Atlas

Code that ran

9 of 9 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 7 have no contract verdict; 2 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/2009.07641.json · badge: /api/ran/2009.07641.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
xxcheng0708/BSNPlusPlus-boundary-sensitive-network community (classifier label)
classifier basis not recorded on the link
9 ran · 0 no recorded run
FunctionStatusWhere it lives
bi_loss Ran
ran on a synthesized input
xxcheng0708/BSNPlusPlus-boundary-sensitive-network/loss_function.py
code served (permissive licence) · get_code("2a1b5cbf26a5e2ba")
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 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

torch.manual_seed(0)

# Create small tensors for pred_score and gt_label
# pred_score should be predictions (floats between 0 and 1)
# gt_label should be binary labels (0 or 1, or values > 0.5 for positive)
pred_score = torch.tensor([[0.2, 0.8, 0.3, 0.9],
                           [0.1, 0.7, 0.6, 0.4]], dtype=torch.float32)

gt_label = torch.tensor([[0.0, 1.0, 0.0, 1.0],
                         [0.0, 1.0, 1.0, 0.0]], dtype=torch.float32)

ARGS = {
    "pred_score": pred_score,
    "gt_label": gt_label
}
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 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
cbg_loss_func Ran
ran on a synthesized input
xxcheng0708/BSNPlusPlus-boundary-sensitive-network/loss_function.py
code served (permissive licence) · get_code("d57bf724ccd7a70d")
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 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

torch.manual_seed(0)

# Create small tensors for the function parameters
# pred_start and pred_end should be probability-like values (0-1 range)
# gt_start and gt_end should be binary labels (0 or 1)

batch_size = 2
seq_len = 4

pred_start = torch.rand(batch_size, seq_len)
pred_end = torch.rand(batch_size, seq_len)
gt_start = torch.randint(0, 2, (batch_size, seq_len)).float()
gt_end = torch.randint(0, 2, (batch_size, seq_len)).float()

ARGS = {
    "pred_start": pred_start,
    "pred_end": pred_end,
    "gt_start": gt_start,
    "gt_end": gt_end
}
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 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
conv_block Ran
ran on a synthesized input
xxcheng0708/BSNPlusPlus-boundary-sensitive-network/models.py
code served (permissive licence) · get_code("83a6fac7349fd1f0")
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 Sequential. 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 = {
    'in_ch': 3,
    'out_ch': 64,
    'kernel_size': 3,
    'stride': 1,
    'bn_layer': True,
    'activate': True,
}
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 Sequential — a Sequential output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Sequential output the recorded ndim 1 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
getDatasetDict Ran
ran on a synthesized input
xxcheng0708/BSNPlusPlus-boundary-sensitive-network/post_processing.py
code served (permissive licence) · get_code("13551e06a372ae9c")
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 pandas as pd
import json
import tempfile
import os

# Create temporary files for CSV and JSON
temp_dir = tempfile.mkdtemp()
csv_file = os.path.join(temp_dir, 'video_info.csv')
json_file = os.path.join(temp_dir, 'video_anno.json')

# Create sample CSV data with a validation subset entry
csv_data = pd.DataFrame({
    'video': ['video_001', 'video_002'],
    'subset': ['validation', 'train']
})
csv_data.to_csv(csv_file, index=False)

# Create sample JSON annotation data
json_data = {
    'video_001': {
        'duration_frame': 150,
        'duration_second': 5.0,
        'feature_frame': 100,
        'annotations': [
            {'segment': [0, 2], 'label': 'action1'},
            {'segment': [2, 4], 'label': 'action2'}
        ]
    },
    'video_002': {
        'duration_frame': 200,
        'duration_second': 8.0,
        'feature_frame': 120,
        'annotations': [
            {'segment': [1, 3], 'label': 'action3'}
        ]
    }
}

with open(json_file, 'w') as f:
    json.dump(json_data, f)

# Define ARGS dict
ARGS = {
    'opt': {
        'video_info': csv_file,
        'video_anno': json_file
    }
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pandas — 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
get_mask Ran
ran on a synthesized input
xxcheng0708/BSNPlusPlus-boundary-sensitive-network/loss_function.py
code served (permissive licence) · get_code("1cd5e9eec5a11010")
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 (4, 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
import numpy as np

torch.manual_seed(0)

# Analyze the function: get_mask takes a single parameter 'tscale' which is used as:
# - An integer to create a [tscale, tscale] numpy array
# - Loop bounds (range(tscale))
# So tscale should be a small integer

tscale = 4

ARGS = {
    "tscale": tscale
}
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 Tensor, ndim 2, dtype family float, shape (4, 4) — shape (4, 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
ioa_with_anchors Ran
ran on a synthesized input
xxcheng0708/BSNPlusPlus-boundary-sensitive-network/utils.py
code served (permissive licence) · get_code("49b3db69edc148a7")
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 (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 numpy as np

# The function ioa_with_anchors takes 4 parameters:
# - anchors_min: array of minimum anchor values
# - anchors_max: array of maximum anchor values
# - box_min: minimum value of the box
# - box_max: maximum value of the box
#
# The function performs element-wise operations on these arrays,
# so they should be numpy arrays with compatible shapes.

# Create small numpy arrays for testing
anchors_min = np.array([0.0, 1.0, 2.0, 3.0])
anchors_max = np.array([2.0, 3.0, 4.0, 5.0])
box_min = np.array([0.5, 1.5, 1.5, 3.5])
box_max = np.array([1.5, 2.5, 3.5, 4.5])

ARGS = {
    'anchors_min': anchors_min,
    'anchors_max': anchors_max,
    'box_min': box_min,
    'box_max': box_max
}
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 (4,) — shape (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
iou_with_anchors Ran
ran on a synthesized input
xxcheng0708/BSNPlusPlus-boundary-sensitive-network/utils.py
code served (permissive licence) · get_code("1ac2acb4c84fccca")
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 (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 numpy as np

# Construct valid arguments for iou_with_anchors
# The function computes Jaccard/IoU scores between anchors and a box
# All parameters are 1D arrays representing intervals (min and max values)

anchors_min = np.array([0.0, 1.0, 2.0, 3.0])
anchors_max = np.array([1.0, 2.0, 3.0, 4.0])
box_min = np.array(0.5)
box_max = np.array(2.5)

ARGS = {
    "anchors_min": anchors_min,
    "anchors_max": anchors_max,
    "box_min": box_min,
    "box_max": box_max,
}
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 (4,) — shape (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
load_json Ran
ran; our contract draft was wrong, not the code
xxcheng0708/BSNPlusPlus-boundary-sensitive-network/dataset.py
code served (permissive licence) · get_code("1e0bb35d27bdc115")
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 tempfile
import os

# Create a temporary JSON file with sample data
temp_dir = tempfile.mkdtemp()
json_file_path = os.path.join(temp_dir, "test.json")

sample_data = {
    "key1": "value1",
    "key2": [1, 2, 3],
    "key3": {"nested": "object"}
}

with open(json_file_path, 'w') as f:
    json.dump(sample_data, f)

ARGS = {
    "file": json_file_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 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_json Ran
ran; our contract draft was wrong, not the code
xxcheng0708/BSNPlusPlus-boundary-sensitive-network/post_processing.py
code served (permissive licence) · get_code("e924a491a8d3290b")
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 tempfile
import os

# Create a temporary JSON file with sample data
temp_dir = tempfile.mkdtemp()
json_file_path = os.path.join(temp_dir, "test_data.json")

sample_data = {
    "key1": "value1",
    "key2": [1, 2, 3],
    "key3": {"nested": "object"}
}

with open(json_file_path, 'w') as f:
    json.dump(sample_data, f)

ARGS = {
    "file": json_file_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 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

Also present in other papers' code

Of 9 samples lifted from this paper's linked repositories, 9 carry a function identity and 0 do not. Of the 9 distinct functions with an identity, 6 are also present in other papers whose code we hold (6 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
load_json 3 other papers
1 dated earlier, 2 later
1907.09702 (2019-07)
BMN: Boundary-Matching Network for Temporal Action Proposal Generation
2204.01680 (2022-04)
TALLFormer: Temporal Action Localization with a Long-memory Transformer
2408.06437 (2024-08)
HAT: History-Augmented Anchor Transformer for Online Temporal Action Localization
ioa_with_anchors 2 other papers
2 dated earlier
1907.09702 (2019-07)
BMN: Boundary-Matching Network for Temporal Action Proposal Generation
1911.11462 (2019-11)
G-TAD: Sub-Graph Localization for Temporal Action Detection
iou_with_anchors 2 other papers
2 dated earlier
1907.09702 (2019-07)
BMN: Boundary-Matching Network for Temporal Action Proposal Generation
1911.11462 (2019-11)
G-TAD: Sub-Graph Localization for Temporal Action Detection
load_json 2 other papers
2 dated earlier
1907.09702 (2019-07)
BMN: Boundary-Matching Network for Temporal Action Proposal Generation
1911.11462 (2019-11)
G-TAD: Sub-Graph Localization for Temporal Action Detection
getDatasetDict 1 other paper
1 dated earlier
1907.09702 (2019-07)
BMN: Boundary-Matching Network for Temporal Action Proposal Generation
get_mask 1 other paper
1 dated earlier
1907.09702 (2019-07)
BMN: Boundary-Matching Network for Temporal Action Proposal Generation

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

Generating human action proposals in untrimmed videos is an important yet challenging task with wide applications. Current methods often suffer from the noisy boundary locations and the inferior quality of confidence scores used for proposal retrieving. In this paper, we present BSN++, a new framework which exploits complementary boundary regressor and relation modeling for temporal proposal generation. First, we propose a novel boundary regressor based on the complementary characteristics of both starting and ending boundary classifiers. Specifically, we utilize the U-shaped architecture with nested skip connections to capture rich contexts and introduce bi-directional boundary matching mechanism to improve boundary precision. Second, to account for the proposal-proposal relations ignored in previous methods, we devise a proposal relation block to which includes two self-attention modules from the aspects of position and channel. Furthermore, we find that there inevitably exists data imbalanced problems in the positive/negative proposals and temporal durations, which harm the model performance on tail distributions. To relieve this issue, we introduce the scale-balanced re-sampling strategy. Extensive experiments are conducted on two popular benchmarks: ActivityNet-1.3 and THUMOS14, which demonstrate that BSN++ achieves the state-of-the-art performance. Not surprisingly, the proposed BSN++ ranked 1st place in the CVPR19 - ActivityNet challenge leaderboard on temporal action localization task.

For agents

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

get_harvested_code_for_paper("2009.07641")
get_code_for_paper("2009.07641")
have("2009.07641")

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

curl https://syntology.ai/api/ran/2009.07641.json

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

[![Syntology run record](https://syntology.ai/api/ran/2009.07641.svg)](https://syntology.ai/paper/2009.07641)

Connect an agent — have() is free.