SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1611.04849 · 2016

Deeply supervised salient object detection with short connections

arXiv · PDF · Open in the Atlas

Code that ran

3 of 3 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 3 have no contract verdict; 0 failures of our instrument; 0 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/1611.04849.json · badge: /api/ran/1611.04849.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
AceCoooool/DSS-pytorch community (classifier label)
classifier basis not recorded on the link
3 ran · 0 no recorded run
FunctionStatusWhere it lives
extra_layer Ran
ran on a synthesized input
AceCoooool/DSS-pytorch/dssnet.py
code served (permissive licence) · get_code("427fee6e3984a2ab")
How it ran

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

import torch
from torch import nn

torch.manual_seed(0)

def vgg(cfg, i=3, batch_norm=False):
    layers = []
    in_channels = i
    for v in cfg:
        if v == 'M':
            layers += [nn.MaxPool2d(kernel_size=2, stride=2)]
        else:
            conv2d = nn.Conv2d(in_channels, v, kernel_size=3, padding=1)
            if batch_norm:
                layers += [conv2d, nn.BatchNorm2d(v), nn.ReLU(inplace=True)]
            else:
                layers += [conv2d, nn.ReLU(inplace=True)]
            in_channels = v
    return layers

class ConcatLayer(nn.Module):

    def __init__(self, list_k, k, scale=True):
        super(ConcatLayer, self).__init__()
        l, up, self.scale = (len(list_k), [], scale)
        for i in range(l):
            up.append(nn.ConvTranspose2d(1, 1, list_k[i], list_k[i] // 2, list_k[i] // 4))
        self.upconv = nn.ModuleList(up)
        self.conv = nn.Conv2d(l + 1, 1, 1, 1)
        self.deconv = nn.ConvTranspose2d(1, 1, k * 2, k, k // 2) if scale else None

    def forward(self, x, list_x):
        elem_x = [x]
        for i, elem in enumerate(list_x):
            elem_x.append(self.upconv[i](elem))
        if self.scale:
            out = self.deconv(self.conv(torch.cat(elem_x, dim=1)))
        else:
            out = self.conv(torch.cat(elem_x, dim=1))
        return out

class FeatLayer(nn.Module):

    def __init__(self, in_channel, channel, k):
        super(FeatLayer, self).__init__()
        self.main = nn.Sequential(nn.Conv2d(in_channel, channel, k, 1, k // 2), nn.ReLU(inplace=True), nn.Conv2d(channel, channel, k, 1, k // 2), nn.ReLU(inplace=True), nn.Conv2d(channel, 1, 1, 1))

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

def extra_layer(vgg, cfg):
    feat_layers, concat_layers, scale = ([], [], 1)
    for k, v in enumerate(cfg):
        feat_layers += [FeatLayer(v[0], v[1], v[2])]
        concat_layers += [ConcatLayer(v[3], scale, k != 0)]
        scale *= 2
    return (vgg, feat_layers, concat_layers)

vgg_base = nn.Sequential(*vgg([64, 'M', 128], i=3, batch_norm=False))

cfg_list = [
    [16, 32, 3, [8, 16]],
    [32, 64, 3, [16, 32]],
    [64, 128, 3, [32, 64]]
]

ARGS = {
    "vgg": vgg_base,
    "cfg": cfg_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 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
get_loader Ran
ran on a synthesized input
AceCoooool/DSS-pytorch/dataset.py
code served (permissive licence) · get_code("c30bbe8adf8ad847")
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 DataLoader. 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 shutil
import torch
from PIL import Image
from torchvision import transforms

torch.manual_seed(0)

# Create temporary directories with minimal test data
temp_dir = tempfile.mkdtemp()
img_root = os.path.join(temp_dir, 'images')
label_root = os.path.join(temp_dir, 'labels')
os.makedirs(img_root)
os.makedirs(label_root)

# Create minimal test images (2 images)
for i in range(2):
    # Create a small RGB image
    img = Image.new('RGB', (32, 32), color='red')
    img.save(os.path.join(img_root, f'image_{i}.jpg'))
    
    # Create a small grayscale label
    label = Image.new('L', (32, 32), color=128)
    label.save(os.path.join(label_root, f'image_{i}.png'))

ARGS = {
    'img_root': img_root,
    'label_root': label_root,
    'img_size': 32,
    'batch_size': 2,
    'filename': None,
    'mode': 'train',
    'num_thread': 0,
    'pin': False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, torch, torchvision — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type DataLoader — a DataLoader output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a DataLoader 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
vgg Ran
ran on a synthesized input
AceCoooool/DSS-pytorch/dssnet.py
code served (permissive licence) · get_code("7ac464b8add50a62")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 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
from torch import nn

torch.manual_seed(0)

# The function vgg takes:
# - cfg: a list/iterable containing integers and 'M' strings
# - i: an integer (input channels, default 3)
# - batch_norm: a boolean (default False)

# Reading the function body:
# - cfg is iterated over, checking if v == 'M' or treating v as an integer (number of output channels)
# - i is used as the initial in_channels value
# - batch_norm is used in an if statement to decide layer composition

# Construct a simple config: mix of channel numbers and max pool markers
cfg_value = [64, 'M', 128, 'M', 256]

ARGS = {
    'cfg': cfg_value,
    'i': 3,
    'batch_norm': 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 list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 3 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

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

Abstract

Recent progress on saliency detection is substantial, benefiting mostly from the explosive development of Convolutional Neural Networks (CNNs). Semantic segmentation and saliency detection algorithms developed lately have been mostly based on Fully Convolutional Neural Networks (FCNs). There is still a large room for improvement over the generic FCN models that do not explicitly deal with the scale-space problem. Holistically-Nested Edge Detector (HED) provides a skip-layer structure with deep supervision for edge and boundary detection, but the performance gain of HED on salience detection is not obvious. In this paper, we propose a new method for saliency detection by introducing short connections to the skip-layer structures within the HED architecture. Our framework provides rich multi-scale feature maps at each layer, a property that is critically needed to perform segment detection. Our method produces state-of-the-art results on 5 widely tested salient object detection benchmarks, with advantages in terms of efficiency (0.15 seconds per image), effectiveness, and simplicity over the existing algorithms.

For agents

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

get_harvested_code_for_paper("1611.04849")
get_code_for_paper("1611.04849")
have("1611.04849")

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

curl https://syntology.ai/api/ran/1611.04849.json

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

[![Syntology run record](https://syntology.ai/api/ran/1611.04849.svg)](https://syntology.ai/paper/1611.04849)

Connect an agent — have() is free.