SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1802.02611 · ECCV · 2018

Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation

Yukun Zhu, Hartwig Adam, Liang-Chieh Chen, Florian Schroff, George Papandreou

arXiv · PDF · Open in the Atlas

Code that ran

44 of 72 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (17 of them constructed an object rather than computing a result; 2 honoured the contract we drafted, 0 violated it and 26 have no contract verdict; 16 failures of our instrument; 28 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/1802.02611.json · badge: /api/ran/1802.02611.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
Vujas-Eteph/CiVOS community
Papers with Code archive
10 ran · 7 no recorded run
petko-nikolov/pysemseg community
not in the paper's declared repositories
7 ran · 2 no recorded run
sara-nl/SURF-segmentation community
not in the paper's declared repositories
5 ran · 2 no recorded run
mlearing/pytorch-deeplab-v3-plus community
not in the paper's declared repositories
4 ran · 1 no recorded run
ensta-u2is/deeplabv3plus-muad-pytorch community
Papers with Code archive
4 ran · 0 no recorded run
samson6460/tf2_Segmentation community
not in the paper's declared repositories
3 ran · 2 no recorded run
giovanniguidi/deeplabV3-PyTorch community
Papers with Code archive
2 ran · 5 no recorded run
JanMarcelKezmann/TensorFlow-Advanced-Segmentation-Models community
Papers with Code archive
2 ran · 2 no recorded run
lattice-ai/DeepLabV3-Plus community
not in the paper's declared repositories
2 ran · 1 no recorded run
Burf/tfdetection community
not in the paper's declared repositories
2 ran · 0 no recorded run
dootmaan/dsrl community
Papers with Code archive
1 ran · 1 no recorded run
MukhlasAdib/KITTI_Mapping community
Papers with Code archive
1 ran · 1 no recorded run
SkyWa7ch3r/SceneSegmentation community
Papers with Code archive
1 ran · 0 no recorded run
kaichoulyc/course_segmentation community
not in the paper's declared repositories
0 ran · 1 no recorded run
jfzhang95/pytorch-deeplab-xception community
not in the paper's declared repositories
0 ran · 1 no recorded run
PuchatekwSzortach/voc_encoder_decoder_with_atrous_separable_convolutions community
Papers with Code archive
0 ran · 1 no recorded run
samson6460/tf2_pose_estimation community
not in the paper's declared repositories
0 ran · 1 no recorded run
FunctionStatusWhere it lives
ASPP Ran
ran on a synthesized input · constructed an object; no result was computed
ensta-u2is/deeplabv3plus-muad-pytorch/network/_deeplab.py
code served (permissive licence) · get_code("9a5496117125b307")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, ASPP: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
from torch import nn

torch.manual_seed(0)

ARGS = {
    "in_channels": 64,
    "atrous_rates": (6, 12, 18),
}
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 ASPP — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a ASPP 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
ASPP Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
SkyWa7ch3r/SceneSegmentation/deeplabv3plus.py
pointer only (licence: NONE) · get_code("0f94e0514fb25366")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, SkyWa7ch3r/SceneSegmentation 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.

ASPPConv Ran
ran on a synthesized input · constructed an object; no result was computed
ensta-u2is/deeplabv3plus-muad-pytorch/network/_deeplab.py
code served (permissive licence) · get_code("eaa0ac6ed86c450b")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, ASPPConv: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
from torch import nn

torch.manual_seed(0)

ARGS = {
    'in_channels': 3,
    'out_channels': 4,
    'dilation': 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 ASPPConv — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a ASPPConv 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
ASPPModule Ran
ran on a synthesized input · constructed an object; no result was computed
petko-nikolov/pysemseg/pysemseg/models/deeplab.py
pointer only (licence: GPL-3.0) · get_code("0633b0a6570b756e")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, petko-nikolov/pysemseg, is not in the permissive set Syntology serves inline. 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.

ASPPPooling Ran
ran on a synthesized input · constructed an object; no result was computed
ensta-u2is/deeplabv3plus-muad-pytorch/network/_deeplab.py
code served (permissive licence) · get_code("cb6b44750325dc1e")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, ASPPPooling: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
from torch import nn

torch.manual_seed(0)

ARGS = {
    'in_channels': 64,
    'out_channels': 256,
}
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 ASPPPooling — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a ASPPPooling 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
ASPP_module Ran
ran on a synthesized input · constructed an object; no result was computed
mlearing/pytorch-deeplab-v3-plus/networks/deeplab_resnet.py
code served (permissive licence) · get_code("a4f0a021c8fd4295")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, ASPP_module: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# ASPP_module constructor takes: inplanes, planes, rate
# inplanes: number of input channels
# planes: number of output channels
# rate: dilation rate (affects kernel_size and padding)

# The forward method expects input x to be a 4D tensor (batch, channels, height, width)
# We need to test with different rate values to ensure the module works correctly

ARGS = {
    "inplanes": 64,      # number of input channels
    "planes": 32,        # number of output channels
    "rate": 2            # dilation rate (3 options: 1, 2, or higher values)
}
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 ASPP_module — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a ASPP_module 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
AtrousSeparableConvolutionBnReLU Ran
ran on a synthesized input · constructed an object; no result was computed
JanMarcelKezmann/TensorFlow-Advanced-Segmentation-Models/tensorflow_advanced_segmentation_models/models/DeepLabV3plus.py
pointer only (licence: NOASSERTION) · get_code("03d31a466787b052")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, JanMarcelKezmann/TensorFlow-Advanced-Segmentation-Models (NOASSERTION). 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.

AtrousSpatialPyramidPooling Ran
ran on a synthesized input · constructed an object; no result was computed
Burf/tfdetection/tfdet/model/head/deeplab.py
code served (permissive licence) · get_code("a2f1b4aadd626233")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, AtrousSpatialPyramidPooling: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import tensorflow as tf

def conv(filters, kernel_size, strides=1, padding='same', use_bias=True, kernel_initializer='he_normal', **kwargs):
    return tf.keras.layers.Conv2D(filters, kernel_size, strides=strides, padding=padding, use_bias=use_bias, kernel_initializer=kernel_initializer, **kwargs)

def normalize(axis=-1, momentum=0.9, epsilon=1e-05, **kwargs):
    return tf.keras.layers.BatchNormalization(axis=axis, momentum=momentum, epsilon=epsilon, **kwargs)

ARGS = {
    'rate': [12, 24, 36],
    'n_feature': 512,
    'method': 'bilinear',
    'convolution': conv,
    'normalize': normalize,
    'activation': tf.keras.activations.relu,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type AtrousSpatialPyramidPooling — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a AtrousSpatialPyramidPooling 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
AtrousSpatialPyramidPooling Ran
ran on a synthesized input · constructed an object; no result was computed
lattice-ai/DeepLabV3-Plus/deeplabv3plus/model/deeplabv3_plus.py
pointer only (licence: NONE) · get_code("4a9a634b49dda7a4")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, lattice-ai/DeepLabV3-Plus 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.

BasicBlockV1b Ran
ran on a synthesized input
Vujas-Eteph/CiVOS/RiTM/isegm/model/modeling/deeplab_v3.py
code served (permissive licence) · get_code("1e87570d52719052")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 4, 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

# Create a downsample module for when stride != 1 or inplanes != planes
inplanes = 8
planes = 16
stride = 2
dilation = 1
previous_dilation = 1

downsample = nn.Sequential(
    nn.Conv2d(inplanes, planes, kernel_size=1, stride=stride, bias=False),
    nn.BatchNorm2d(planes)
)

INIT_ARGS = {
    'inplanes': inplanes,
    'planes': planes,
    'stride': stride,
    'dilation': dilation,
    'downsample': downsample,
    'previous_dilation': previous_dilation,
    'norm_layer': nn.BatchNorm2d
}

# Create input tensor with shape (batch, channels, height, width)
x = torch.randn(2, inplanes, 12, 10)

ARGS = {
    'x': x
}
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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 — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
Bottleneck Ran
ran on a synthesized input · constructed an object; no result was computed
mlearing/pytorch-deeplab-v3-plus/networks/deeplab_resnet.py
code served (permissive licence) · get_code("682f6d56a1c45f69")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, Bottleneck: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# The Bottleneck module expects:
# - inplanes: number of input channels
# - planes: number of output channels (before expansion)
# - stride: stride for conv2
# - rate: dilation rate for conv2
# - downsample: optional nn.Module to match dimensions

# Create a simple downsample module that matches the residual to output dimensions
# The output of conv3 has planes*4 channels, so downsample must map inplanes to planes*4
downsample_module = nn.Sequential(
    nn.Conv2d(16, 64, kernel_size=1, stride=2, bias=False),
    nn.BatchNorm2d(64)
)

ARGS = {
    "inplanes": 16,
    "planes": 16,
    "stride": 2,
    "rate": 1,
    "downsample": downsample_module
}
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 Bottleneck — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Bottleneck 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
Bottleneck Ran
ran on a synthesized input · constructed an object; no result was computed
petko-nikolov/pysemseg/pysemseg/models/deeplab.py
pointer only (licence: GPL-3.0) · get_code("df2ed20fe3290ced")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, petko-nikolov/pysemseg, is not in the permissive set Syntology serves inline. 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.

BottleneckV1b Ran
ran on a synthesized input
Vujas-Eteph/CiVOS/RiTM/isegm/model/modeling/deeplab_v3.py
code served (permissive licence) · get_code("43562ef9d87c54e7")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 4, dtype family float, shape (2, 64, 8, 8) 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
from torch import nn

class BottleneckV1b(nn.Module):
    expansion = 4

    def __init__(self, inplanes, planes, stride=1, dilation=1, downsample=None, previous_dilation=1, norm_layer=nn.BatchNorm2d):
        super(BottleneckV1b, self).__init__()
        self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=1, bias=False)
        self.bn1 = norm_layer(planes)
        self.conv2 = nn.Conv2d(planes, planes, kernel_size=3, stride=stride, padding=dilation, dilation=dilation, bias=False)
        self.bn2 = norm_layer(planes)
        self.conv3 = nn.Conv2d(planes, planes * self.expansion, kernel_size=1, bias=False)
        self.bn3 = norm_layer(planes * self.expansion)
        self.relu = nn.ReLU(inplace=True)
        self.downsample = downsample
        self.stride = stride

    def forward(self, x):
        residual = x
        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)
        out = self.conv2(out)
        out = self.bn2(out)
        out = self.relu(out)
        out = self.conv3(out)
        out = self.bn3(out)
        if self.downsample is not None:
            residual = self.downsample(x)
        out = out + residual
        out = self.relu(out)
        return out


# Create a simple downsample module for the residual connection
downsample_module = nn.Sequential(
    nn.Conv2d(16, 64, kernel_size=1, stride=1, bias=False),
    nn.BatchNorm2d(64)
)

INIT_ARGS = {
    'inplanes': 16,
    'planes': 16,
    'stride': 1,
    'dilation': 1,
    'downsample': downsample_module,
    'previous_dilation': 1,
    'norm_layer': nn.BatchNorm2d
}

ARGS = {
    'x': torch.randn(2, 16, 8, 8)
}

if __name__ == '__main__':
    model = BottleneckV1b(**INIT_ARGS)
    model.eval()
    output = model(**ARGS)
    print(f"Output shape: {output.shape}")
    print(f"Expected shape: torch.Size([2, 64, 8, 8])")
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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, 64, 8, 8) — shape (2, 64, 8, 8) 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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
ConvBlock Ran
ran on a synthesized input
lattice-ai/DeepLabV3-Plus/deeplabv3plus/model/deeplabv3_plus.py
pointer only (licence: NONE) · get_code("692e7f2e56fbd23e")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, lattice-ai/DeepLabV3-Plus 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.

ConvolutionBnActivation Ran
ran on a synthesized input · constructed an object; no result was computed
JanMarcelKezmann/TensorFlow-Advanced-Segmentation-Models/tensorflow_advanced_segmentation_models/models/DeepLabV3plus.py
pointer only (licence: NOASSERTION) · get_code("3326d9cd4c357d28")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, JanMarcelKezmann/TensorFlow-Advanced-Segmentation-Models (NOASSERTION). 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.

Decoder Ran
ran on a synthesized input · constructed an object; no result was computed
petko-nikolov/pysemseg/pysemseg/models/deeplab.py
pointer only (licence: GPL-3.0) · get_code("303dec08f3ebd517")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, petko-nikolov/pysemseg, is not in the permissive set Syntology serves inline. 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.

DeepLabHeadV3Plus Ran
ran on a synthesized input · constructed an object; no result was computed
ensta-u2is/deeplabv3plus-muad-pytorch/network/_deeplab.py
code served (permissive licence) · get_code("fc4e40cedd008787")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, DeepLabHeadV3Plus: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn
from torch.nn import functional as F

torch.manual_seed(0)

# Define the helper classes that are used by DeepLabHeadV3Plus
class ASPPConv(nn.Sequential):
    def __init__(self, in_channels, out_channels, dilation):
        modules = [nn.Conv2d(in_channels, out_channels, 3, padding=dilation, dilation=dilation, bias=False), nn.BatchNorm2d(out_channels), nn.ReLU(inplace=True)]
        super(ASPPConv, self).__init__(*modules)

class ASPPPooling(nn.Sequential):
    def __init__(self, in_channels, out_channels):
        super(ASPPPooling, self).__init__(nn.AdaptiveAvgPool2d(1), nn.Conv2d(in_channels, out_channels, 1, bias=False), nn.BatchNorm2d(out_channels), nn.ReLU(inplace=True))

    def forward(self, x):
        size = x.shape[-2:]
        x = super(ASPPPooling, self).forward(x)
        return F.interpolate(x, size=size, mode='bilinear', align_corners=False)

class ASPP(nn.Module):
    def __init__(self, in_channels, atrous_rates):
        super(ASPP, self).__init__()
        out_channels = 256
        modules = []
        modules.append(nn.Sequential(nn.Conv2d(in_channels, out_channels, 1, bias=False), nn.BatchNorm2d(out_channels), nn.ReLU(inplace=True)))
        rate1, rate2, rate3 = tuple(atrous_rates)
        modules.append(ASPPConv(in_channels, out_channels, rate1))
        modules.append(ASPPConv(in_channels, out_channels, rate2))
        modules.append(ASPPConv(in_channels, out_channels, rate3))
        modules.append(ASPPPooling(in_channels, out_channels))
        self.convs = nn.ModuleList(modules)
        self.project = nn.Sequential(nn.Conv2d(5 * out_channels, out_channels, 1, bias=False), nn.BatchNorm2d(out_channels), nn.ReLU(inplace=True), nn.Dropout(0.1))

    def forward(self, x):
        res = []
        for conv in self.convs:
            res.append(conv(x))
        res = torch.cat(res, dim=1)
        return self.project(res)

class DeepLabHeadV3Plus(nn.Module):
    def __init__(self, in_channels, low_level_channels, num_classes, aspp_dilate=[12, 24, 36]):
        super(DeepLabHeadV3Plus, self).__init__()
        self.project = nn.Sequential(nn.Conv2d(low_level_channels, 48, 1, bias=False), nn.BatchNorm2d(48), nn.ReLU(inplace=True))
        self.aspp = ASPP(in_channels, aspp_dilate)
        self.classifier = nn.Sequential(nn.Conv2d(304, 256, 3, padding=1, bias=False), nn.BatchNorm2d(256), nn.ReLU(inplace=True), nn.Conv2d(256, num_classes, 1))
        self._init_weight()

    def forward(self, feature):
        low_level_feature = self.project(feature['low_level'])
        output_feature = self.aspp(feature['out'])
        output_feature = F.interpolate(output_feature, size=low_level_feature.shape[2:], mode='bilinear', align_corners=False)
        return self.classifier(torch.cat([low_level_feature, output_feature], dim=1))

    def _init_weight(self):
        for m in self.modules():
            if isinstance(m, nn.Conv2d):
                nn.init.kaiming_normal_(m.weight)
            elif isinstance(m, (nn.BatchNorm2d, nn.GroupNorm)):
                nn.init.constant_(m.weight, 1)
                nn.init.constant_(m.bias, 0)

ARGS = {
    'in_channels': 512,
    'low_level_channels': 64,
    'num_classes': 21,
    'aspp_dilate': [12, 24, 36]
}
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 DeepLabHeadV3Plus — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a DeepLabHeadV3Plus 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
Deeplab Ran
ran on a synthesized input · constructed an object; no result was computed
petko-nikolov/pysemseg/pysemseg/models/deeplab.py
pointer only (licence: GPL-3.0) · get_code("c10a37b7673d30e0")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, petko-nikolov/pysemseg, is not in the permissive set Syntology serves inline. 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.

ResBlock Ran
ran on a synthesized input · constructed an object; no result was computed
petko-nikolov/pysemseg/pysemseg/models/deeplab.py
pointer only (licence: GPL-3.0) · get_code("d99dfd9a3f26d56b")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, petko-nikolov/pysemseg, is not in the permissive set Syntology serves inline. 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.

ResNet Ran
ran on a synthesized input · constructed an object; no result was computed
mlearing/pytorch-deeplab-v3-plus/networks/deeplab_resnet.py
code served (permissive licence) · get_code("42610493d6f82abe")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, ResNet: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# Define a minimal block class that satisfies the usage pattern
class MockBlock(nn.Module):
    expansion = 1
    
    def __init__(self, inplanes, planes, stride=1, rate=1, downsample=None):
        super(MockBlock, self).__init__()
        self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=3, stride=stride, padding=1, bias=False)
        self.bn1 = nn.BatchNorm2d(planes)
        self.relu = nn.ReLU(inplace=True)
        self.downsample = downsample
        self.stride = stride
    
    def forward(self, x):
        identity = x
        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)
        if self.downsample is not None:
            identity = self.downsample(x)
        out += identity
        return out

ARGS = {
    'nInputChannels': 3,
    'block': MockBlock,
    'layers': [3, 4, 6],
    'os': 16,
    'pretrained': False
}
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 ResNet — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a ResNet 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
ResNet Ran
ran on a synthesized input · constructed an object; no result was computed
petko-nikolov/pysemseg/pysemseg/models/deeplab.py
pointer only (licence: GPL-3.0) · get_code("8e495c6bb6122711")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, petko-nikolov/pysemseg, is not in the permissive set Syntology serves inline. 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.

ResNet101 Ran
ran; our contract draft was wrong, not the code
mlearing/pytorch-deeplab-v3-plus/networks/deeplab_resnet.py
code served (permissive licence) · get_code("5b3058ed46740fef")
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 ResNet. 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 = {
    "nInputChannels": 3,
    "os": 16,
    "pretrained": False,
}
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 ResNet — a ResNet output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ResNet 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
SepConv_BN Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
sara-nl/SURF-segmentation/deeplab/keras-deeplab-v3-plus-master/model.py
pointer only (licence: NONE) · get_code("b293f1d81655c5e6")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, sara-nl/SURF-segmentation 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.

SepConv_BN Ran
ran; our contract draft was wrong, not the code
samson6460/tf2_Segmentation/models/deep_lab_model.py
pointer only (licence: NONE) · get_code("e6cccdb5cc9404d7")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, samson6460/tf2_Segmentation 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.

SeparableConv2d Ran
ran on a synthesized input
Vujas-Eteph/CiVOS/RiTM/isegm/model/modeling/deeplab_v3.py
code served (permissive licence) · get_code("414adf534687df82")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 4, 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
from torch import nn
import types as _types

def select_activation_function(activation):
    if isinstance(activation, str):
        if activation.lower() == 'relu':
            return nn.ReLU
        elif activation.lower() == 'softplus':
            return nn.Softplus
        else:
            raise ValueError(f'Unknown activation type {activation}')
    elif isinstance(activation, nn.Module):
        return activation
    else:
        raise ValueError(f'Unknown activation type {activation}')

class SeparableConv2d(nn.Module):

    def __init__(self, in_channels, out_channels, dw_kernel, dw_padding, dw_stride=1, activation=None, use_bias=False, norm_layer=None):
        super(SeparableConv2d, self).__init__()
        _activation = select_activation_function(activation)
        self.body = nn.Sequential(
            nn.Conv2d(in_channels, in_channels, kernel_size=dw_kernel, stride=dw_stride, padding=dw_padding, bias=use_bias, groups=in_channels),
            nn.Conv2d(in_channels, out_channels, kernel_size=1, stride=1, bias=use_bias),
            norm_layer(out_channels) if norm_layer is not None else nn.Identity(),
            _activation()
        )

    def forward(self, x):
        return self.body(x)

INIT_ARGS = {
    'in_channels': 3,
    'out_channels': 4,
    'dw_kernel': 3,
    'dw_padding': 1,
    'dw_stride': 1,
    'activation': 'relu',
    'use_bias': False,
    'norm_layer': nn.BatchNorm2d
}

ARGS = {
    'x': torch.randn(2, 3, 5, 6)
}
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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 — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
_ASPP Ran
ran on a synthesized input
Vujas-Eteph/CiVOS/RiTM/isegm/model/modeling/deeplab_v3.py
code served (permissive licence) · get_code("ebedb218ce19918d")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 4, dtype family float, shape (2, 32, 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
from torch import nn
import torch.nn.functional as F

def _ASPPConv(in_channels, out_channels, atrous_rate, norm_layer):
    block = nn.Sequential(nn.Conv2d(in_channels=in_channels, out_channels=out_channels, kernel_size=3, padding=atrous_rate, dilation=atrous_rate, bias=False), norm_layer(out_channels), nn.ReLU())
    return block

class _AsppPooling(nn.Module):

    def __init__(self, in_channels, out_channels, norm_layer):
        super(_AsppPooling, self).__init__()
        self.gap = nn.Sequential(nn.AdaptiveAvgPool2d((1, 1)), nn.Conv2d(in_channels=in_channels, out_channels=out_channels, kernel_size=1, bias=False), norm_layer(out_channels), nn.ReLU())

    def forward(self, x):
        pool = self.gap(x)
        return F.interpolate(pool, x.size()[2:], mode='bilinear', align_corners=True)

class _ASPP(nn.Module):

    def __init__(self, in_channels, atrous_rates, out_channels=256, project_dropout=0.5, norm_layer=nn.BatchNorm2d):
        super(_ASPP, self).__init__()
        b0 = nn.Sequential(nn.Conv2d(in_channels=in_channels, out_channels=out_channels, kernel_size=1, bias=False), norm_layer(out_channels), nn.ReLU())
        rate1, rate2, rate3 = tuple(atrous_rates)
        b1 = _ASPPConv(in_channels, out_channels, rate1, norm_layer)
        b2 = _ASPPConv(in_channels, out_channels, rate2, norm_layer)
        b3 = _ASPPConv(in_channels, out_channels, rate3, norm_layer)
        b4 = _AsppPooling(in_channels, out_channels, norm_layer=norm_layer)
        self.concurent = nn.ModuleList([b0, b1, b2, b3, b4])
        project = [nn.Conv2d(in_channels=5 * out_channels, out_channels=out_channels, kernel_size=1, bias=False), norm_layer(out_channels), nn.ReLU()]
        if project_dropout > 0:
            project.append(nn.Dropout(project_dropout))
        self.project = nn.Sequential(*project)

    def forward(self, x):
        x = torch.cat([block(x) for block in self.concurent], dim=1)
        return self.project(x)


INIT_ARGS = {
    'in_channels': 8,
    'atrous_rates': (6, 12, 18),
    'out_channels': 32,
    'project_dropout': 0.5,
    'norm_layer': nn.BatchNorm2d
}

ARGS = {
    'x': torch.randn(2, 8, 4, 3)
}

if __name__ == '__main__':
    model = _ASPP(**INIT_ARGS)
    model.eval()
    output = model(**ARGS)
    print(f"Output shape: {output.shape}")
    print(f"Expected shape: (2, 32, 4, 3)")
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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, 32, 4, 3) — shape (2, 32, 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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
_ASPPConv Ran
ran; our contract draft was wrong, not the code
Vujas-Eteph/CiVOS/RiTM/isegm/model/modeling/deeplab_v3.py
code served (permissive licence) · get_code("3274a200ef8fa29a")
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 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
from torch import nn

torch.manual_seed(0)

# Define the arguments for _ASPPConv
in_channels = 3
out_channels = 64
atrous_rate = 2
norm_layer = nn.BatchNorm2d

ARGS = {
    "in_channels": in_channels,
    "out_channels": out_channels,
    "atrous_rate": atrous_rate,
    "norm_layer": norm_layer,
}
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 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
_AsppPooling Ran
ran on a synthesized input
Vujas-Eteph/CiVOS/RiTM/isegm/model/modeling/deeplab_v3.py
code served (permissive licence) · get_code("d5caae9d47ff2394")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 4, dtype family float, shape (2, 8, 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 torch.nn as nn
from torch import nn as nn_module

INIT_ARGS = {
    'in_channels': 16,
    'out_channels': 8,
    'norm_layer': nn.BatchNorm2d,
}

ARGS = {
    'x': torch.randn(2, 16, 4, 3),
}

if __name__ == '__main__':
    from torch import nn
    import torch.nn.functional as F

    class _AsppPooling(nn.Module):

        def __init__(self, in_channels, out_channels, norm_layer):
            super(_AsppPooling, self).__init__()
            self.gap = nn.Sequential(nn.AdaptiveAvgPool2d((1, 1)), nn.Conv2d(in_channels=in_channels, out_channels=out_channels, kernel_size=1, bias=False), norm_layer(out_channels), nn.ReLU())

        def forward(self, x):
            pool = self.gap(x)
            return F.interpolate(pool, x.size()[2:], mode='bilinear', align_corners=True)

    module = _AsppPooling(**INIT_ARGS)
    output = module(**ARGS)
    print(f"Output shape: {output.shape}")
    print(f"Expected shape: {ARGS['x'].shape}")
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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, 8, 4, 3) — shape (2, 8, 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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
_DeepLabHead Ran
ran on a synthesized input
Vujas-Eteph/CiVOS/RiTM/isegm/model/modeling/deeplab_v3.py
code served (permissive licence) · get_code("a5651a6111de513f")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 4, 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
from torch import nn
import types as _types

def select_activation_function(activation):
    if isinstance(activation, str):
        if activation.lower() == 'relu':
            return nn.ReLU
        elif activation.lower() == 'softplus':
            return nn.Softplus
        else:
            raise ValueError(f'Unknown activation type {activation}')
    elif isinstance(activation, nn.Module):
        return activation
    else:
        raise ValueError(f'Unknown activation type {activation}')

class SeparableConv2d(nn.Module):

    def __init__(self, in_channels, out_channels, dw_kernel, dw_padding, dw_stride=1, activation=None, use_bias=False, norm_layer=None):
        super(SeparableConv2d, self).__init__()
        _activation = ops.select_activation_function(activation)
        self.body = nn.Sequential(nn.Conv2d(in_channels, in_channels, kernel_size=dw_kernel, stride=dw_stride, padding=dw_padding, bias=use_bias, groups=in_channels), nn.Conv2d(in_channels, out_channels, kernel_size=1, stride=1, bias=use_bias), norm_layer(out_channels) if norm_layer is not None else nn.Identity(), _activation())

    def forward(self, x):
        return self.body(x)

class _DeepLabHead(nn.Module):

    def __init__(self, out_channels, in_channels, mid_channels=256, norm_layer=nn.BatchNorm2d):
        super(_DeepLabHead, self).__init__()
        self.block = nn.Sequential(SeparableConv2d(in_channels=in_channels, out_channels=mid_channels, dw_kernel=3, dw_padding=1, activation='relu', norm_layer=norm_layer), SeparableConv2d(in_channels=mid_channels, out_channels=mid_channels, dw_kernel=3, dw_padding=1, activation='relu', norm_layer=norm_layer), nn.Conv2d(in_channels=mid_channels, out_channels=out_channels, kernel_size=1))

    def forward(self, x):
        return self.block(x)

ops = _types.SimpleNamespace(**{n: globals()[n] for n in ['select_activation_function'] if n in globals()})

INIT_ARGS = {
    'out_channels': 8,
    'in_channels': 4,
    'mid_channels': 16,
    'norm_layer': nn.BatchNorm2d
}

ARGS = {
    'x': torch.randn(2, 4, 5, 6)
}

if __name__ == '__main__':
    model = _DeepLabHead(**INIT_ARGS)
    output = model(**ARGS)
    print(f"Output shape: {output.shape}")
    print(f"Expected shape: (2, 8, 5, 6)")
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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 — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
_SkipProject Ran
ran on a synthesized input
Vujas-Eteph/CiVOS/RiTM/isegm/model/modeling/deeplab_v3.py
code served (permissive licence) · get_code("9c74246baf89e0d2")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 4, dtype family float, shape (2, 4, 5, 6) 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
from torch import nn
import types as _types

def select_activation_function(activation):
    if isinstance(activation, str):
        if activation.lower() == 'relu':
            return nn.ReLU
        elif activation.lower() == 'softplus':
            return nn.Softplus
        else:
            raise ValueError(f'Unknown activation type {activation}')
    elif isinstance(activation, nn.Module):
        return activation
    else:
        raise ValueError(f'Unknown activation type {activation}')

class _SkipProject(nn.Module):

    def __init__(self, in_channels, out_channels, norm_layer=nn.BatchNorm2d):
        super(_SkipProject, self).__init__()
        _activation = ops.select_activation_function('relu')
        self.skip_project = nn.Sequential(nn.Conv2d(in_channels, out_channels, kernel_size=1, bias=False), norm_layer(out_channels), _activation())

    def forward(self, x):
        return self.skip_project(x)


ops = _types.SimpleNamespace(**{n: globals()[n] for n in ['select_activation_function'] if n in globals()})

INIT_ARGS = {
    'in_channels': 3,
    'out_channels': 4,
    'norm_layer': nn.BatchNorm2d
}

ARGS = {
    'x': torch.randn(2, 3, 5, 6)
}

if __name__ == '__main__':
    model = _SkipProject(**INIT_ARGS)
    output = model(**ARGS)
    print(f"Input shape: {ARGS['x'].shape}")
    print(f"Output shape: {output.shape}")
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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, 4, 5, 6) — shape (2, 4, 5, 6) 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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
_conv2d_same Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
sara-nl/SURF-segmentation/deeplab/keras-deeplab-v3-plus-master/model.py
pointer only (licence: NONE) · get_code("3ebb19f692e505f6")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, sara-nl/SURF-segmentation 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.

_conv2d_same Ran
ran; our contract draft was wrong, not the code
samson6460/tf2_Segmentation/models/deep_lab_model.py
pointer only (licence: NONE) · get_code("1096e5ba9c24632c")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, samson6460/tf2_Segmentation 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.

_make_divisible Ran
ran, honoured the contract we drafted
sara-nl/SURF-segmentation/deeplab/keras-deeplab-v3-plus-master/model.py
pointer only (licence: NONE) · get_code("1eac80b6e2cee7b2")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, sara-nl/SURF-segmentation 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.

_safe_state_dict_filtering Ran
ran; our contract draft was wrong, not the code
Vujas-Eteph/CiVOS/RiTM/isegm/model/modeling/deeplab_v3.py
code served (permissive licence) · get_code("5776a2cbccc2bd40")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned dict. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# Create a sample original state dict (simulating a loaded checkpoint)
orig_dict = {
    'layer1.weight': torch.randn(4, 3),
    'layer1.bias': torch.randn(4),
    'layer2.weight': torch.randn(5, 4),
    'unused_param': torch.randn(2, 2),  # This won't be in model_dict_keys
}

# Create a set of keys that the model actually has
model_dict_keys = {
    'layer1.weight',
    'layer1.bias',
    'layer2.weight',
}

ARGS = {
    'orig_dict': orig_dict,
    'model_dict_keys': model_dict_keys,
}
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
_xception_block Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
sara-nl/SURF-segmentation/deeplab/keras-deeplab-v3-plus-master/model.py
pointer only (licence: NONE) · get_code("76185251accca7d2")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, sara-nl/SURF-segmentation 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.

_xception_block Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
samson6460/tf2_Segmentation/models/deep_lab_model.py
pointer only (licence: NONE) · get_code("398cd429c73d683e")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, samson6460/tf2_Segmentation 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.

build_aspp Ran
ran; our contract draft was wrong, not the code
giovanniguidi/deeplabV3-PyTorch/models/deeplab.py
pointer only (licence: NONE) · get_code("68330450cf875f58")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, giovanniguidi/deeplabV3-PyTorch 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.

build_decoder Ran
ran; our contract draft was wrong, not the code
giovanniguidi/deeplabV3-PyTorch/models/deeplab.py
pointer only (licence: NONE) · get_code("3f8ff8379eaf97df")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, giovanniguidi/deeplabV3-PyTorch 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.

build_sr_decoder Ran
ran; our contract draft was wrong, not the code
dootmaan/dsrl/modeling/deeplab.py
code served (permissive licence) · get_code("6d5672a93099e8b5")
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 Decoder. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

class SynchronizedBatchNorm2d(nn.Module):
    def __init__(self, num_features):
        super(SynchronizedBatchNorm2d, self).__init__()
        self.weight = nn.Parameter(torch.ones(num_features))
        self.bias = nn.Parameter(torch.zeros(num_features))
    
    def forward(self, x):
        return x

class Decoder(nn.Module):
    def __init__(self, num_classes, backbone, BatchNorm):
        super(Decoder, self).__init__()
        if backbone == 'resnet' or backbone == 'drn':
            low_level_inplanes = 256
        elif backbone == 'xception':
            low_level_inplanes = 128
        elif backbone == 'mobilenet':
            low_level_inplanes = 24
        else:
            raise NotImplementedError
        self.conv1 = nn.Conv2d(low_level_inplanes, 48, 1, bias=False)
        self.bn1 = BatchNorm(48)
        self.relu = nn.ReLU()
        self.last_conv = nn.Sequential(
            nn.Conv2d(304, 256, kernel_size=3, stride=1, padding=1, bias=False),
            BatchNorm(256),
            nn.ReLU(),
            nn.Dropout(0.5),
            nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1, bias=False),
            BatchNorm(256),
            nn.ReLU(),
            nn.Dropout(0.1),
            nn.Conv2d(256, num_classes, kernel_size=1, stride=1)
        )
        self._init_weight()

    def forward(self, x, low_level_feat):
        low_level_feat = self.conv1(low_level_feat)
        low_level_feat = self.bn1(low_level_feat)
        low_level_feat = self.relu(low_level_feat)
        x = torch.nn.functional.interpolate(x, size=low_level_feat.size()[2:], mode='bilinear', align_corners=True)
        x = torch.cat((x, low_level_feat), dim=1)
        x = self.last_conv(x)
        return x

    def _init_weight(self):
        for m in self.modules():
            if isinstance(m, nn.Conv2d):
                torch.nn.init.kaiming_normal_(m.weight)
            elif isinstance(m, SynchronizedBatchNorm2d):
                m.weight.data.fill_(1)
                m.bias.data.zero_()
            elif isinstance(m, nn.BatchNorm2d):
                m.weight.data.fill_(1)
                m.bias.data.zero_()

num_classes = 19
backbone = 'resnet'
BatchNorm = nn.BatchNorm2d

decoder = Decoder(num_classes, backbone, BatchNorm)

x = torch.randn(2, 256, 4, 5)
low_level_feat = torch.randn(2, 256, 8, 10)

ARGS = {
    'num_classes': 19,
    'backbone': 'resnet',
    'BatchNorm': nn.BatchNorm2d
}
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 Decoder — a Decoder output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Decoder 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
conv Ran
ran; our contract draft was wrong, not the code
Burf/tfdetection/tfdet/model/head/deeplab.py
code served (permissive licence) · get_code("b387bf7cf1bc8ced")
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 Conv2D. 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 tensorflow as tf

ARGS = {
    'filters': 32,
    'kernel_size': 3,
    'strides': 1,
    'padding': 'same',
    'use_bias': True,
    'kernel_initializer': 'he_normal',
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Conv2D — a Conv2D output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Conv2D 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
relu6 Ran
ran, honoured the contract we drafted
sara-nl/SURF-segmentation/deeplab/keras-deeplab-v3-plus-master/model.py
pointer only (licence: NONE) · get_code("e7a210769e38826c")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, sara-nl/SURF-segmentation 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.

resnet Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
petko-nikolov/pysemseg/pysemseg/models/deeplab.py
pointer only (licence: GPL-3.0) · get_code("90c8dc399d2c6b51")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, petko-nikolov/pysemseg, is not in the permissive set Syntology serves inline. 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.

segmentation_inference Ran
ran on a synthesized input
MukhlasAdib/KITTI_Mapping/app/pipeline/road_segmentation.py
pointer only (licence: NONE) · get_code("b1992cb92fab6136")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, MukhlasAdib/KITTI_Mapping 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_activation_function Ran
ran; our contract draft was wrong, not the code
Vujas-Eteph/CiVOS/RiTM/isegm/model/modeling/deeplab_v3.py
code served (permissive licence) · get_code("4976637b084e5281")
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 type. 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 torch import nn

torch.manual_seed(0)

# Test case 1: string activation 'relu'
activation_str_relu = 'relu'

# Test case 2: string activation 'softplus'
activation_str_softplus = 'softplus'

# Test case 3: nn.Module instance
activation_module = nn.ReLU()

# Create ARGS dict with a string activation (testing 'relu')
ARGS = {
    'activation': activation_str_relu
}
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 type — a type output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a type 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
ASPP Not yet run
unverified (harvested, no recorded run)
giovanniguidi/deeplabV3-PyTorch/models/deeplab.py
pointer only (licence: NONE) · get_code("26047a3c90df0a36")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, giovanniguidi/deeplabV3-PyTorch 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.

AtrousSpatialPyramidPoolingV3 Not yet run
unverified (harvested, no recorded run)
JanMarcelKezmann/TensorFlow-Advanced-Segmentation-Models/tensorflow_advanced_segmentation_models/models/DeepLabV3plus.py
pointer only (licence: NOASSERTION) · get_code("a9f549942de7d30b")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, JanMarcelKezmann/TensorFlow-Advanced-Segmentation-Models (NOASSERTION). 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.

Decoder Not yet run
unverified (harvested, no recorded run)
giovanniguidi/deeplabV3-PyTorch/models/deeplab.py
pointer only (licence: NONE) · get_code("edf673cbd75a7027")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, giovanniguidi/deeplabV3-PyTorch 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.

DeepLab Not yet run
unverified (harvested, no recorded run)
giovanniguidi/deeplabV3-PyTorch/models/deeplab.py
pointer only (licence: NONE) · get_code("4ac4bc7c6f72419c")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, giovanniguidi/deeplabV3-PyTorch 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.

DeepLab Not yet run
unverified (harvested, no recorded run)
kaichoulyc/course_segmentation/model/deep_lab/deeplab.py
pointer only (licence: NONE) · get_code("6e6115c04c3f121d")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, kaichoulyc/course_segmentation 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.

DeepLab Not yet run
unverified (harvested, no recorded run)
dootmaan/dsrl/modeling/deeplab.py
code served (permissive licence) · get_code("9ac514a69bb5bb62")
How it ran

no invocation recorded.

DeepLab Not yet run
unverified (harvested, no recorded run)
jfzhang95/pytorch-deeplab-xception/modeling/deeplab.py
code served (permissive licence) · get_code("09a240d3865fee6c")
How it ran

no invocation recorded.

DeepLabV3Plus Not yet run
unverified (harvested, no recorded run)
Vujas-Eteph/CiVOS/RiTM/isegm/model/modeling/deeplab_v3.py
code served (permissive licence) · get_code("609cfe0df52cf002")
How it ran

no invocation recorded.

DeepLabV3PlusBuilder Not yet run
unverified (harvested, no recorded run)
PuchatekwSzortach/voc_encoder_decoder_with_atrous_separable_convolutions/net/ml.py
code served (permissive licence) · get_code("9755f047bfec69ce")
How it ran

no invocation recorded.

DeepLabV3plus Not yet run
unverified (harvested, no recorded run)
JanMarcelKezmann/TensorFlow-Advanced-Segmentation-Models/tensorflow_advanced_segmentation_models/models/DeepLabV3plus.py
pointer only (licence: NOASSERTION) · get_code("63a49a3c1dd4d263")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, JanMarcelKezmann/TensorFlow-Advanced-Segmentation-Models (NOASSERTION). 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.

DeepLabv3_plus Not yet run
unverified (harvested, no recorded run)
mlearing/pytorch-deeplab-v3-plus/networks/deeplab_resnet.py
code served (permissive licence) · get_code("dfd28692d87464d1")
How it ran

no invocation recorded.

DeeplabV3Plus Not yet run
unverified (harvested, no recorded run)
lattice-ai/DeepLabV3-Plus/deeplabv3plus/model/deeplabv3_plus.py
pointer only (licence: NONE) · get_code("59849f10377fad5f")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, lattice-ai/DeepLabV3-Plus 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.

Deeplabv3 Not yet run
unverified (harvested, no recorded run)
sara-nl/SURF-segmentation/deeplab/keras-deeplab-v3-plus-master/model.py
pointer only (licence: NONE) · get_code("13a67fcfc1e4c1d7")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, sara-nl/SURF-segmentation 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.

ResNetBackbone Not yet run
unverified (harvested, no recorded run)
Vujas-Eteph/CiVOS/RiTM/isegm/model/modeling/deeplab_v3.py
code served (permissive licence) · get_code("fabf7df16100c5bc")
How it ran

no invocation recorded.

ResNetV1b Not yet run
unverified (harvested, no recorded run)
Vujas-Eteph/CiVOS/RiTM/isegm/model/modeling/deeplab_v3.py
code served (permissive licence) · get_code("96875ada06fa67da")
How it ran

no invocation recorded.

RoadSegmentation Not yet run
unverified (harvested, no recorded run)
MukhlasAdib/KITTI_Mapping/app/pipeline/road_segmentation.py
pointer only (licence: NONE) · get_code("82f03ca51ba7dbc4")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, MukhlasAdib/KITTI_Mapping 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.

Showing the first 60. An agent gets every one from get_harvested_code_for_paper("1802.02611").

Also present in other papers' code

Of 72 samples lifted from this paper's linked repositories, 72 carry a function identity and 0 do not. Of the 70 distinct functions with an identity, 14 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 14 later) and 56 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
_make_divisible
brick of this paper's assembly: helper
11 other papers
2 dated earlier, 9 later · 7 only as an assembly brick
1704.04861 (2017-04)
MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
1801.04381 (2018-01)
MobileNetV2: Inverted Residuals and Linear Bottlenecks
1905.02244 (2019-05)
Searching for MobileNetV3
and 8 more not named here
ASPP
brick of this paper's assembly: model
6 other papers
6 later · 3 only as an assembly brick
2103.06533 (2021-03)
Triple-cooperative Video Shadow Detection
2106.11562 (2021-06)
SSUL: Semantic Segmentation with Unknown Label for Exemplar-based Class-Incremental Learning
2205.08083 (2022-05)
Region-Aware Metric Learning for Open World Semantic Segmentation via Meta-Channel Aggregation
and 3 more not named here
ASPPConv
brick of this paper's assembly: helper
6 other papers
6 later · 5 only as an assembly brick
2103.06533 (2021-03)
Triple-cooperative Video Shadow Detection
2106.11562 (2021-06)
SSUL: Semantic Segmentation with Unknown Label for Exemplar-based Class-Incremental Learning
2205.08083 (2022-05)
Region-Aware Metric Learning for Open World Semantic Segmentation via Meta-Channel Aggregation
and 3 more not named here
ASPPPooling
brick of this paper's assembly: helper
6 other papers
6 later · 4 only as an assembly brick
2103.06533 (2021-03)
Triple-cooperative Video Shadow Detection
2106.11562 (2021-06)
SSUL: Semantic Segmentation with Unknown Label for Exemplar-based Class-Incremental Learning
2205.08083 (2022-05)
Region-Aware Metric Learning for Open World Semantic Segmentation via Meta-Channel Aggregation
and 3 more not named here
build_aspp
brick of this paper's assembly: helper
6 other papers
6 later
1910.02550 (2019-10)
ClearGrasp: 3D Shape Estimation of Transparent Objects for Manipulation
2010.06349 (2020-10)
Collaborative Video Object Segmentation by Multi-Scale Foreground-Background Integration
2104.06394 (2021-04)
All you need are a few pixels: semantic segmentation with PixelPick
and 3 more not named here
ASPP
brick of this paper's assembly: entry, model
4 other papers
4 later · 2 only as an assembly brick
2105.10886 (2021-05)
VS-Net: Voting with Segmentation for Visual Localization
2108.00384 (2021-08)
Edge-competing Pathological Liver Vessel Segmentation with Limited Labels
2109.01801 (2021-09)
Dual Transfer Learning for Event-based End-task Prediction via Pluggable Event to Image Translation
and 1 more not named here
_ASPPModule
brick of this paper's assembly: helper
4 other papers
4 later · 2 only as an assembly brick
2105.10886 (2021-05)
VS-Net: Voting with Segmentation for Visual Localization
2108.00384 (2021-08)
Edge-competing Pathological Liver Vessel Segmentation with Limited Labels
2109.01801 (2021-09)
Dual Transfer Learning for Event-based End-task Prediction via Pluggable Event to Image Translation
and 1 more not named here
DeepLabHeadV3Plus
brick of this paper's assembly: entry
3 other papers
3 later · 1 only as an assembly brick
2404.07155 (2024-04)
Unified Language-driven Zero-shot Domain Adaptation
2603.20021 (2026-03)
ODySSeI: An Open-Source End-to-End Framework for Automated Detection, Segmentation, and Severity Estimation of Lesions in Invasive Coronary Angiography Images
2604.18256 (2026-04)
Domain-Specialized Object Detection via Model-Level Mixtures of Experts
build_decoder
brick of this paper's assembly: helper
2 other papers
2 later
1910.02550 (2019-10)
ClearGrasp: 3D Shape Estimation of Transparent Objects for Manipulation
2108.00384 (2021-08)
Edge-competing Pathological Liver Vessel Segmentation with Limited Labels
BasicBlockV1b
brick of this paper's assembly: helper
1 other paper
1 later
2204.02574 (2022-04)
FocalClick: Towards Practical Interactive Image Segmentation
BottleneckV1b
brick of this paper's assembly: helper
1 other paper
1 later
2204.02574 (2022-04)
FocalClick: Towards Practical Interactive Image Segmentation
Decoder
brick of this paper's assembly: model
1 other paper
1 later
2108.00384 (2021-08)
Edge-competing Pathological Liver Vessel Segmentation with Limited Labels
build_backbone
brick of this paper's assembly: helper
1 other paper
1 later · 1 only as an assembly brick
2108.00384 (2021-08)
Edge-competing Pathological Liver Vessel Segmentation with Limited Labels
select_activation_function
brick of this paper's assembly: helper
1 other paper
1 later
2404.00741 (2024-04)
Rethinking Interactive Image Segmentation with Low Latency, High Quality, and Diverse Prompts

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

More are linked than shown here.

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

Abstract

Spatial pyramid pooling module or encode-decoder structure are used in deep neural networks for semantic segmentation task. The former networks are able to encode multi-scale contextual information by probing the incoming features with filters or pooling operations at multiple rates and multiple effective fields-of-view, while the latter networks can capture sharper object boundaries by gradually recovering the spatial information. In this work, we propose to combine the advantages from both methods. Specifically, our proposed model, DeepLabv3+, extends DeepLabv3 by adding a simple yet effective decoder module to refine the segmentation results especially along object boundaries. We further explore the Xception model and apply the depthwise separable convolution to both Atrous Spatial Pyramid Pooling and decoder modules, resulting in a faster and stronger encoder-decoder network. We demonstrate the effectiveness of the proposed model on PASCAL VOC 2012 and Cityscapes datasets, achieving the test set performance of 89.0% and 82.1% without any post-processing. Our paper is accompanied with a publicly available reference implementation of the proposed models in Tensorflow at https: //github.com/tensorflow/models/tree/master/research/deeplab.

For agents

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

get_harvested_code_for_paper("1802.02611")
get_code_for_paper("1802.02611")
have("1802.02611")

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

curl https://syntology.ai/api/ran/1802.02611.json

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

[![Syntology run record](https://syntology.ai/api/ran/1802.02611.svg)](https://syntology.ai/paper/1802.02611)

Connect an agent — have() is free.