SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2005.09007 · 2020

U$^2$-Net: Going Deeper with Nested U-Structure for Salient Object Detection

arXiv · PDF · Open in the Atlas

Code that ran

24 of 42 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; 1 honoured the contract we drafted, 1 violated it and 20 have no contract verdict; 2 failures of our instrument; 18 no recorded run), measured 2026-11-21; 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/2005.09007.json · badge: /api/ran/2005.09007.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
NathanUA/U-2-Net canonical (classifier label)
classifier basis: census
1 ran · 1 no recorded run
xuebinqin/U-2-Net canonical (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
RoadoneP/aloha community (classifier label)
classifier basis not recorded on the link
8 ran · 1 no recorded run
nadermx/backgroundremover community (classifier label)
classifier basis not recorded on the link
4 ran · 1 no recorded run
singhaman092/rembg community (classifier label)
classifier basis not recorded on the link
3 ran · 4 no recorded run
Akhilesh64/Image-Segmentation-U-2-Net community (classifier label)
classifier basis not recorded on the link
2 ran · 3 no recorded run
levindabhi/cloth-segmentation community (classifier label)
classifier basis not recorded on the link
2 ran · 2 no recorded run
sk1123344/U-2-NET-and-U-NET reimplementation (classifier label)
classifier basis: classifier_2025
1 ran · 1 no recorded run
patharanordev/background-remover community (classifier label)
classifier basis not recorded on the link
1 ran · 1 no recorded run
hhaAndroid/u2net reimplementation (classifier label)
classifier basis: classifier_2025
1 ran · 1 no recorded run
yingkaisha/keras-unet-collection reimplementation (classifier label)
classifier basis: classifier_2025
0 ran · 3 no recorded run
FunctionStatusWhere it lives
bce_loss_all Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
sk1123344/U-2-NET-and-U-NET/u2net_train_test.py
pointer only (licence: NONE) · get_code("81a853f579032ae4")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, sk1123344/U-2-NET-and-U-NET 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.

center_size Ran
ran on a synthesized input
RoadoneP/aloha/layers/box_utils.py
code served (permissive licence) · get_code("0118ac3b2ad5818b")
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; 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

torch.manual_seed(0)

# The function expects boxes in point form (xmin, ymin, xmax, ymax)
# and converts them to center-size form (cx, cy, w, h)
# boxes should be a tensor of shape (N, 4) where N is number of boxes
boxes = torch.tensor([
    [10.0, 20.0, 30.0, 40.0],
    [5.0, 15.0, 25.0, 35.0]
], dtype=torch.float32)

ARGS = {
    "boxes": boxes
}
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 — no output shape is stated: observe_signature.py could not observe it (import:OSError)
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
check_bidi Ran
ran on a synthesized input
singhaman092/rembg/other/.serverless/requirements/idna/core.py
code served (permissive licence) · get_code("354103a1d87fedfd")
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 unicodedata

class IDNAError(UnicodeError):
    """ Base exception for all IDNA-encoding related problems """
    pass

class IDNABidiError(IDNAError):
    """ Exception when bidirectional requirements are not satisfied """
    pass

def check_bidi(label: str, check_ltr: bool=False) -> bool:
    bidi_label = False
    for idx, cp in enumerate(label, 1):
        direction = unicodedata.bidirectional(cp)
        if direction == '':
            raise IDNABidiError('Unknown directionality in label {} at position {}'.format(repr(label), idx))
        if direction in ['R', 'AL', 'AN']:
            bidi_label = True
    if not bidi_label and (not check_ltr):
        return True
    direction = unicodedata.bidirectional(label[0])
    if direction in ['R', 'AL']:
        rtl = True
    elif direction == 'L':
        rtl = False
    else:
        raise IDNABidiError('First codepoint in label {} must be directionality L, R or AL'.format(repr(label)))
    valid_ending = False
    number_type = None
    for idx, cp in enumerate(label, 1):
        direction = unicodedata.bidirectional(cp)
        if rtl:
            if not direction in ['R', 'AL', 'AN', 'EN', 'ES', 'CS', 'ET', 'ON', 'BN', 'NSM']:
                raise IDNABidiError('Invalid direction for codepoint at position {} in a right-to-left label'.format(idx))
            if direction in ['R', 'AL', 'EN', 'AN']:
                valid_ending = True
            elif direction != 'NSM':
                valid_ending = False
            if direction in ['AN', 'EN']:
                if not number_type:
                    number_type = direction
                elif number_type != direction:
                    raise IDNABidiError('Can not mix numeral types in a right-to-left label')
        else:
            if not direction in ['L', 'EN', 'ES', 'CS', 'ET', 'ON', 'BN', 'NSM']:
                raise IDNABidiError('Invalid direction for codepoint at position {} in a left-to-right label'.format(idx))
            if direction in ['L', 'EN']:
                valid_ending = True
            elif direction != 'NSM':
                valid_ending = False
    if not valid_ending:
        raise IDNABidiError('Label ends with illegal codepoint directionality')
    return True

ARGS = {
    'label': 'hello',
    'check_ltr': 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 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
construct_backbone Ran
ran on a synthesized input
RoadoneP/aloha/backbone.py
code served (permissive licence) · get_code("87523070b15ff05c")
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 SimpleBackbone. 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


class SimpleBackbone(nn.Module):
    def __init__(self, input_dim=4, hidden_dim=8):
        super().__init__()
        self.layers = nn.ModuleList()
        self.input_dim = input_dim
        self.hidden_dim = hidden_dim
        # Initialize with one layer
        self.layers.append(nn.Linear(input_dim, hidden_dim))
    
    def add_layer(self):
        """Add a new layer to the backbone."""
        self.layers.append(nn.Linear(self.hidden_dim, self.hidden_dim))
    
    def forward(self, x):
        for layer in self.layers:
            x = layer(x)
        return x


class BackboneConfig:
    def __init__(self, type_class, args, selected_layers):
        self.type = type_class
        self.args = args
        self.selected_layers = selected_layers


torch.manual_seed(0)

cfg = BackboneConfig(
    type_class=SimpleBackbone,
    args=(4, 8),  # input_dim, hidden_dim
    selected_layers=[0, 1, 2]  # Request layers 0, 1, 2 (so max is 2, need 3 layers total)
)

ARGS = {
    "cfg": cfg
}
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 SimpleBackbone — a SimpleBackbone output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a SimpleBackbone 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
crop_face Ran
ran on a synthesized input
NathanUA/U-2-Net/u2net_portrait_demo.py
code served (permissive licence) · get_code("b42507b856af3374")
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 3, dtype family int, shape (512, 512, 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 cv2
import numpy as np

# Create a sample image (height=480, width=640, 3 channels for BGR)
img = np.random.randint(0, 256, (480, 640, 3), dtype=np.uint8)

# Create a face bounding box as (x, y, w, h)
# x=100, y=80, w=120, h=150
face = (100, 80, 120, 150)

ARGS = {
    "img": img,
    "face": face
}
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 ndarray, ndim 3, dtype family int, shape (512, 512, 3) — shape (512, 512, 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
download_files_from_github Ran
ran on a synthesized input
nadermx/backgroundremover/backgroundremover/github.py
code served (permissive licence) · get_code("b752b6ca8885fd9e")
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 os
import tempfile

# Create a temporary directory for the download path
temp_dir = tempfile.gettempdir()
download_path = os.path.join(temp_dir, "test_model.pth")

ARGS = {
    "path": download_path,
    "model_name": "u2netp",
    "max_retries": 3
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
requests — 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
getData Ran
ran on a synthesized input
Akhilesh64/Image-Segmentation-U-2-Net/dataLoader.py
code served (permissive licence) · get_code("21c24045e17c6985")
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 os
import tempfile
import numpy as np
from PIL import Image

# Create temporary directories with dummy image files
train_dir = tempfile.mkdtemp()
gt_dir = tempfile.mkdtemp()

# Create 2 sample training images (256x256 RGB)
for i in range(2):
    train_img = Image.new('RGB', (256, 256), color=(100 + i*50, 150, 200))
    train_img.save(os.path.join(train_dir, f'train_{i}.png'))
    
    # Create corresponding ground truth image (256x256 grayscale)
    gt_img = Image.new('L', (256, 256), color=(128 + i*30))
    gt_img.save(os.path.join(gt_dir, f'gt_{i}.png'))

ARGS = {
    'train_path': train_dir,
    'gt_path': gt_dir
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy, tensorflow — 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
getData Ran
ran on a synthesized input
Akhilesh64/Image-Segmentation-U-2-Net/Scripts_DUTS/dataLoader.py
code served (permissive licence) · get_code("8144398e708b679f")
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 os
import tempfile
import numpy as np
from PIL import Image

# Create temporary directories with test images
train_dir = tempfile.mkdtemp()
gt_dir = tempfile.mkdtemp()

# Create dummy training images (256x256 RGB)
for i in range(10):
    img = Image.new('RGB', (256, 256), color=(100, 150, 200))
    img.save(os.path.join(train_dir, f'train_{i:03d}.png'))

# Create dummy ground truth images (256x256 grayscale)
for i in range(10):
    img = Image.new('L', (256, 256), color=128)
    img.save(os.path.join(gt_dir, f'gt_{i:03d}.png'))

ARGS = {
    'train_path': train_dir,
    'gt_path': gt_dir,
    'epoch': 0,
    'batch_size': 2
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy, tensorflow — 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
gram Ran
ran on a synthesized input
RoadoneP/aloha/style_utils.py
code served (permissive licence) · get_code("93ebd13b9e693ca3")
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 3, dtype family float, shape (2, 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

torch.manual_seed(0)

# The function takes a tensor and computes its Gram matrix
# It expects shape (B, C, H, W) where:
# B = batch size
# C = channels
# H = height
# W = width

# Create a small test tensor with distinct dimensions
# B=2, C=4, H=3, W=5 (all different for easy shape probing)
tensor = torch.randn(2, 4, 3, 5)

ARGS = {
    "tensor": tensor
}
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 3, dtype family float, shape (2, 4, 4) — shape (2, 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
intersect Ran
ran on a synthesized input
RoadoneP/aloha/layers/box_utils.py
code served (permissive licence) · get_code("2b050648fe0422de")
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 3, 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

torch.manual_seed(0)

# Create bounding box tensors
# Shape requirements: [n, A, 4] and [n, B, 4]
# where n=batch size, A=number of boxes in first set, B=number of boxes in second set
# Each box is [x1, y1, x2, y2]

n = 2  # batch size
A = 3  # number of boxes in box_a
B = 4  # number of boxes in box_b

# box_a: [n, A, 4] = [2, 3, 4]
box_a = torch.tensor([
    [[0.0, 0.0, 10.0, 10.0], [5.0, 5.0, 15.0, 15.0], [20.0, 20.0, 25.0, 25.0]],
    [[1.0, 1.0, 11.0, 11.0], [6.0, 6.0, 16.0, 16.0], [21.0, 21.0, 26.0, 26.0]]
], dtype=torch.float32)

# box_b: [n, B, 4] = [2, 4, 4]
box_b = torch.tensor([
    [[0.0, 0.0, 5.0, 5.0], [8.0, 8.0, 12.0, 12.0], [20.0, 20.0, 23.0, 23.0], [30.0, 30.0, 35.0, 35.0]],
    [[1.0, 1.0, 6.0, 6.0], [9.0, 9.0, 13.0, 13.0], [21.0, 21.0, 24.0, 24.0], [31.0, 31.0, 36.0, 36.0]]
], dtype=torch.float32)

ARGS = {
    "box_a": box_a,
    "box_b": box_b
}
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 3, dtype family float — no output shape is stated: observe_signature.py could not observe it (import:OSError)
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
itot Ran
ran on a synthesized input
RoadoneP/aloha/style_utils.py
code served (permissive licence) · get_code("e38d55db4e9ff568")
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 4, dtype family float, shape (1, 3, 4, 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 torch
import numpy as np
from torchvision import transforms

torch.manual_seed(0)

# The function expects:
# - img: an image that can be processed by either ToTensor or ToPILImage+ToTensor
# - max_size: None or some value (used to determine which transform pipeline)

# Create a simple numpy array that can be converted to a PIL image or tensor
# PIL images expect uint8 arrays with shape (H, W, C)
img_array = np.random.randint(0, 256, (4, 3, 3), dtype=np.uint8)

ARGS = {
    "img": img_array,
    "max_size": None,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch, torchvision — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 4, dtype family float, shape (1, 3, 4, 3) — shape (1, 3, 4, 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
load_image Ran
ran on a synthesized input
RoadoneP/aloha/style_utils.py
code served (permissive licence) · get_code("4969a06ac02c02d2")
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 3, dtype family int, shape (10, 10, 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 cv2
import numpy as np
import tempfile
import os

# Create a temporary image file for testing
temp_dir = tempfile.mkdtemp()
temp_image_path = os.path.join(temp_dir, "test_image.png")

# Create a small test image using numpy and save it
test_image = np.random.randint(0, 256, (10, 10, 3), dtype=np.uint8)
cv2.imwrite(temp_image_path, test_image)

ARGS = {
    "path": temp_image_path
}
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 ndarray, ndim 3, dtype family int, shape (10, 10, 3) — shape (10, 10, 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
max_workers Ran
ran on a synthesized input
nadermx/backgroundremover/backgroundremover/bg.py
code served (permissive licence) · get_code("f4893b2a0f0fc410")
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. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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 torch.nn as nn
import os

torch.manual_seed(0)

ARGS = {
    'model_name': 'u2net',
    'gpu_batchsize': 2,
}
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
muti_bce_loss_fusion Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
xuebinqin/U-2-Net/u2net_train.py
code served (permissive licence) · get_code("fc74645243d46d51")
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) and returned tuple, 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 torch
import torch.nn as nn

torch.manual_seed(0)

# Create small binary tensors for predictions (output of model layers)
# BCELoss expects input in range [0, 1], so we use sigmoid outputs
batch_size = 2
num_classes = 4

d0 = torch.sigmoid(torch.randn(batch_size, num_classes))
d1 = torch.sigmoid(torch.randn(batch_size, num_classes))
d2 = torch.sigmoid(torch.randn(batch_size, num_classes))
d3 = torch.sigmoid(torch.randn(batch_size, num_classes))
d4 = torch.sigmoid(torch.randn(batch_size, num_classes))
d5 = torch.sigmoid(torch.randn(batch_size, num_classes))
d6 = torch.sigmoid(torch.randn(batch_size, num_classes))

# Create binary labels tensor matching the shape
labels_v = torch.randint(0, 2, (batch_size, num_classes)).float()

ARGS = {
    'd0': d0,
    'd1': d1,
    'd2': d2,
    'd3': d3,
    'd4': d4,
    'd5': d5,
    'd6': d6,
    'labels_v': labels_v
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family float — a tuple 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
naive_cutout Ran
ran on a synthesized input
nadermx/backgroundremover/backgroundremover/bg.py
code served (permissive licence) · get_code("f250811a24b78ae9")
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 Image. 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.

from PIL import Image
import numpy as np

# Create a small RGB image
img = Image.new('RGB', (64, 48), color=(255, 0, 0))

# Create a small mask image (grayscale)
mask = Image.new('L', (32, 24), color=255)

ARGS = {
    'img': img,
    'mask': mask,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Image — a Image output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Image output the recorded ndim 3 and dtype family int 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
naive_cutout Ran
ran on a synthesized input
patharanordev/background-remover/src/backgroundremover/bg.py
code served (permissive licence) · get_code("e3da1e070910201f")
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 Image. 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
from PIL import Image
import numpy as np

torch.manual_seed(0)

# Create a small RGB image
img = Image.new('RGB', (4, 3), color=(255, 0, 0))

# Create a small mask image (grayscale) - same or different size, will be resized
mask = Image.new('L', (2, 2), color=128)

ARGS = {
    'img': img,
    'mask': mask
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Image — a Image output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Image output the recorded ndim 3 and dtype family int 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
normPRED Ran
ran, honoured the contract we drafted
hhaAndroid/u2net/u2net_test.py
code served (permissive licence) · get_code("89e4412c100b0f9b")
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) 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)

# Create a small tensor with distinct dimensions
d = torch.randn(2, 4)

ARGS = {
    "d": d
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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
norm_pred Ran
ran on a synthesized input
nadermx/backgroundremover/backgroundremover/u2net/detect.py
code served (permissive licence) · get_code("cfded26c9f3e75e8")
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)

# Create a small tensor with distinct values
d = torch.tensor([[1.0, 2.0, 3.0, 4.0],
                   [5.0, 6.0, 7.0, 8.0]])

ARGS = {
    "d": d
}
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
point_form Ran
ran on a synthesized input
RoadoneP/aloha/layers/box_utils.py
code served (permissive licence) · get_code("27dbfe70566c2e97")
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; 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

torch.manual_seed(0)

# The function takes boxes in center-size format (cx, cy, w, h)
# and converts to corner format (xmin, ymin, xmax, ymax)
# boxes shape: (N, 4) where each row is [cx, cy, w, h]

boxes = torch.tensor([
    [10.0, 20.0, 4.0, 6.0],
    [30.0, 40.0, 8.0, 12.0]
], dtype=torch.float32)

ARGS = {
    "boxes": boxes
}
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 — no output shape is stated: observe_signature.py could not observe it (import:OSError)
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
str2bool Ran
ran, violated the contract we drafted
RoadoneP/aloha/eval_image.py
code served (permissive licence) · get_code("f017532fc389cbfe")
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) 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 str2bool(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.')

# Setup code to construct valid arguments
ARGS = {
    'v': 'true'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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
tensor_for_board Ran
ran on a synthesized input
levindabhi/cloth-segmentation/utils/tensorboard_utils.py
code served (permissive licence) · get_code("30ce730ccf4be37e")
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 4, dtype family float, shape (2, 3, 4, 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 torch

torch.manual_seed(0)

# Create a small image tensor (batch_size=2, channels=1, height=4, width=3)
img_tensor = torch.randn(2, 1, 4, 3)

ARGS = {
    "img_tensor": img_tensor
}
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 4, dtype family float, shape (2, 3, 4, 3) — shape (2, 3, 4, 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
tensor_list_for_board Ran
ran on a synthesized input
levindabhi/cloth-segmentation/utils/tensorboard_utils.py
code served (permissive licence) · get_code("ae4ce56419ea054d")
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 4, dtype family float, shape (2, 3, 8, 15) 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 expects a list of lists of tensors
# Each inner list represents a row in the grid
# Each tensor should be a 4D tensor (batch, channel, height, width)
# The function uses tensor_for_board which expects tensors in range [-1, 1]

# Create some small test tensors
batch_size = 2
channels = 3
height = 4
width = 5

# Create a list of lists of tensors
# First row: 2 tensors
tensor_1 = torch.randn(batch_size, channels, height, width) * 2 - 1  # Range [-1, 1]
tensor_2 = torch.randn(batch_size, channels, height, width) * 2 - 1

# Second row: 3 tensors
tensor_3 = torch.randn(batch_size, channels, height, width) * 2 - 1
tensor_4 = torch.randn(batch_size, channels, height, width) * 2 - 1
tensor_5 = torch.randn(batch_size, channels, height, width) * 2 - 1

img_tensors_list = [
    [tensor_1, tensor_2],
    [tensor_3, tensor_4, tensor_5]
]

ARGS = {
    "img_tensors_list": img_tensors_list
}
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 4, dtype family float, shape (2, 3, 8, 15) — shape (2, 3, 8, 15) 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
valid_label_length Ran
ran on a synthesized input
singhaman092/rembg/other/.serverless/requirements/idna/core.py
code served (permissive licence) · get_code("82202184fa5b7ac9")
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 torch

# Construct arguments for valid_label_length
# The function takes a label parameter that can be either bytes or str
# and checks if its length is <= 63

ARGS = {
    "label": "short_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 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
valid_string_length Ran
ran on a synthesized input
singhaman092/rembg/other/.serverless/requirements/idna/core.py
code served (permissive licence) · get_code("c2eaff651c95d1f4")
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 torch
import numpy as np

torch.manual_seed(0)

# Analyze the function:
# - label: Union[bytes, str] - needs to be either bytes or str
# - trailing_dot: bool - needs to be a boolean
# - The function checks len(label) against thresholds (254 or 253)
# - Returns a boolean based on length validation

# Create test arguments
ARGS = {
    "label": "example.com",
    "trailing_dot": 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 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
REBNCONV Not yet run
unverified (harvested, no recorded run)
Akhilesh64/Image-Segmentation-U-2-Net/U2-net.py
code served (permissive licence) · get_code("1109359fbff5405b")
How it ran

no invocation recorded.

RSU Not yet run
unverified (harvested, no recorded run)
yingkaisha/keras-unet-collection/keras_unet_collection/_model_u2net_2d.py
code served (permissive licence) · get_code("c0d03d6bccec6ee1")
How it ran

no invocation recorded.

RSU4F Not yet run
unverified (harvested, no recorded run)
yingkaisha/keras-unet-collection/keras_unet_collection/_model_u2net_2d.py
code served (permissive licence) · get_code("2e03f0a5f07cb196")
How it ran

no invocation recorded.

RSU7 Not yet run
unverified (harvested, no recorded run)
Akhilesh64/Image-Segmentation-U-2-Net/U2-net.py
code served (permissive licence) · get_code("a28583cb5a89e1f2")
How it ran

no invocation recorded.

darknetconvlayer Not yet run
unverified (harvested, no recorded run)
RoadoneP/aloha/backbone.py
code served (permissive licence) · get_code("901e174e8a74774c")
How it ran

no invocation recorded.

detect_single_face Not yet run
unverified (harvested, no recorded run)
NathanUA/U-2-Net/u2net_portrait_demo.py
code served (permissive licence) · get_code("e5a6c2c113b235f3")
How it ran

no invocation recorded.

get_model Not yet run
unverified (harvested, no recorded run)
patharanordev/background-remover/src/backgroundremover/bg.py
code served (permissive licence) · get_code("8e82437510a87611")
How it ran

no invocation recorded.

get_request Not yet run
unverified (harvested, no recorded run)
singhaman092/rembg/other/utils/utils.py
code served (permissive licence) · get_code("71858fc2cedc576f")
How it ran

no invocation recorded.

load_checkpoint Not yet run
unverified (harvested, no recorded run)
levindabhi/cloth-segmentation/utils/saving_utils.py
code served (permissive licence) · get_code("ae7eca3db4489538")
How it ran

no invocation recorded.

load_checkpoint_mgpu Not yet run
unverified (harvested, no recorded run)
levindabhi/cloth-segmentation/utils/saving_utils.py
code served (permissive licence) · get_code("ec0fd63b77ab92d6")
How it ran

no invocation recorded.

load_model Not yet run
unverified (harvested, no recorded run)
singhaman092/rembg/src/rembg/u2net/detect.py
code served (permissive licence) · get_code("82b06ae41a637f16")
How it ran

no invocation recorded.

loss Not yet run
unverified (harvested, no recorded run)
Akhilesh64/Image-Segmentation-U-2-Net/U2-net.py
code served (permissive licence) · get_code("14100923eb6556ac")
How it ran

no invocation recorded.

muti_bce_loss_fusion Not yet run
unverified (harvested, no recorded run)
hhaAndroid/u2net/u2net_train.py
code served (permissive licence) · get_code("9093b315696852d6")
How it ran

no invocation recorded.

post_request Not yet run
unverified (harvested, no recorded run)
singhaman092/rembg/other/utils/utils.py
code served (permissive licence) · get_code("b13016e1f4154113")
How it ran

no invocation recorded.

pred_norm Not yet run
unverified (harvested, no recorded run)
sk1123344/U-2-NET-and-U-NET/u2net_train_test.py
pointer only (licence: NONE) · get_code("1300ad8439b1c5bf")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, sk1123344/U-2-NET-and-U-NET 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. No invocation recorded.

preprocess Not yet run
unverified (harvested, no recorded run)
nadermx/backgroundremover/backgroundremover/u2net/detect.py
code served (permissive licence) · get_code("c20a71a3b663ddb2")
How it ran

no invocation recorded.

preprocess Not yet run
unverified (harvested, no recorded run)
singhaman092/rembg/src/rembg/u2net/detect.py
code served (permissive licence) · get_code("c576953d98bccacc")
How it ran

no invocation recorded.

u2net_2d_base Not yet run
unverified (harvested, no recorded run)
yingkaisha/keras-unet-collection/keras_unet_collection/_model_u2net_2d.py
code served (permissive licence) · get_code("b6c716b95682bb86")
How it ran

no invocation recorded.

Also present in other papers' code

Of 42 samples lifted from this paper's linked repositories, 42 carry a function identity and 0 do not. Of the 40 distinct functions with an identity, 9 are also present in other papers whose code we hold (8 with at least one paper dated earlier, 1 the same month, 4 later) and 31 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
str2bool 82 other papers
35 dated earlier, 1 the same month, 46 later
1510.03055 (2015-10)
A Diversity-Promoting Objective Function for Neural Conversation Models
1703.06490 (2017-03)
Generating Multi-label Discrete Patient Records using Generative Adversarial Networks
1710.10467 (2017-10)
Generalized End-to-End Loss for Speaker Verification
and 79 more not named here
construct_backbone 4 other papers
3 dated earlier, 1 later
1904.02689 (2019-04)
YOLACT: Real-time Instance Segmentation
1911.09070 (2019-11)
EfficientDet: Scalable and Efficient Object Detection
1912.06218 (2019-12)
YOLACT++: Better Real-time Instance Segmentation
and 1 more not named here
normPRED 4 other papers
4 later
2101.04704 (2021-01)
Boundary-Aware Segmentation Network for Mobile and Web Applications
2308.02097 (2023-08)
Multi-interactive Feature Learning and a Full-time Multi-modality Benchmark for Image Fusion and Segmentation
2411.11045 (2024-11)
StableV2V: Stablizing Shape Consistency in Video-to-Video Editing
and 1 more not named here
darknetconvlayer 3 other papers
3 dated earlier
1904.02689 (2019-04)
YOLACT: Real-time Instance Segmentation
1911.09070 (2019-11)
EfficientDet: Scalable and Efficient Object Detection
1912.06218 (2019-12)
YOLACT++: Better Real-time Instance Segmentation
center_size 2 other papers
2 dated earlier
1904.02689 (2019-04)
YOLACT: Real-time Instance Segmentation
1912.06218 (2019-12)
YOLACT++: Better Real-time Instance Segmentation
intersect 2 other papers
1 dated earlier, 1 later
1912.06218 (2019-12)
YOLACT++: Better Real-time Instance Segmentation
2110.11219 (2021-10)
PlaneRecNet: Multi-Task Learning with Cross-Task Consistency for Piece-Wise Plane Detection and Reconstruction from a Single RGB Image
point_form 2 other papers
2 dated earlier
1904.02689 (2019-04)
YOLACT: Real-time Instance Segmentation
1912.06218 (2019-12)
YOLACT++: Better Real-time Instance Segmentation
tensor_for_board 1 other paper
1 dated earlier
1807.07688 (2018-07)
Toward Characteristic-Preserving Image-based Virtual Try-On Network
tensor_list_for_board 1 other paper
1 dated earlier
1807.07688 (2018-07)
Toward Characteristic-Preserving Image-based Virtual Try-On Network

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 2026-11-21.

Repositories linked to this paper

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

Abstract

In this paper, we design a simple yet powerful deep network architecture, U$^2$-Net, for salient object detection (SOD). The architecture of our U$^2$-Net is a two-level nested U-structure. The design has the following advantages: (1) it is able to capture more contextual information from different scales thanks to the mixture of receptive fields of different sizes in our proposed ReSidual U-blocks (RSU), (2) it increases the depth of the whole architecture without significantly increasing the computational cost because of the pooling operations used in these RSU blocks. This architecture enables us to train a deep network from scratch without using backbones from image classification tasks. We instantiate two models of the proposed architecture, U$^2$-Net (176.3 MB, 30 FPS on GTX 1080Ti GPU) and U$^2$-Net$^{\dagger}$ (4.7 MB, 40 FPS), to facilitate the usage in different environments. Both models achieve competitive performance on six SOD datasets. The code is available: https://github.com/NathanUA/U-2-Net.

For agents

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

get_harvested_code_for_paper("2005.09007")
get_code_for_paper("2005.09007")
have("2005.09007")

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

curl https://syntology.ai/api/ran/2005.09007.json

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

[![Syntology run record](https://syntology.ai/api/ran/2005.09007.svg)](https://syntology.ai/paper/2005.09007)

Connect an agent — have() is free.