SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2203.03041 · ECCV · 2022

Highly Accurate Dichotomous Image Segmentation

Ling Shao, Luc Gool, Xiaobin Hu, Deng-Ping Fan, Xuebin Qin, Hang Dai

arXiv · PDF · Open in the Atlas

Code that ran

9 of 10 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (6 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 6 have no contract verdict; 3 failures of our instrument; 1 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/2203.03041.json · badge: /api/ran/2203.03041.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
xuebinqin/DIS marked official in Papers with Code
Papers with Code archive
9 ran · 1 no recorded run
FunctionStatusWhere it lives
REBNCONV Ran
ran on a synthesized input · constructed an object; no result was computed
xuebinqin/DIS/IS-Net/models/isnet.py
code served (permissive licence) · get_code("afa31b2882bc3c34")
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, REBNCONV: 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)

# Create a tiny input tensor: batch_size=2, channels=3, height=4, width=4
x = torch.randn(2, 3, 4, 4)

ARGS = {
    'in_ch': 3,
    'out_ch': 3,
    'dirate': 1,
    'stride': 1
}
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 REBNCONV — 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 REBNCONV 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
RSU4 Ran
ran on a synthesized input · constructed an object; no result was computed
xuebinqin/DIS/IS-Net/models/isnet.py
code served (permissive licence) · get_code("02cc50421ae7b439")
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, RSU4: 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)

ARGS = {
    'in_ch': 3,
    'mid_ch': 12,
    'out_ch': 3,
}
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 RSU4 — 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 RSU4 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
RSU4F Ran
ran on a synthesized input · constructed an object; no result was computed
xuebinqin/DIS/IS-Net/models/isnet.py
code served (permissive licence) · get_code("b61cd5f9e93344c3")
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, RSU4F: 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)

ARGS = {
    'in_ch': 3,
    'mid_ch': 12,
    'out_ch': 3,
}
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 RSU4F — 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 RSU4F 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
RSU5 Ran
ran on a synthesized input · constructed an object; no result was computed
xuebinqin/DIS/IS-Net/models/isnet.py
code served (permissive licence) · get_code("4e74b24c0513bd83")
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, RSU5: 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)

ARGS = {
    'in_ch': 3,
    'mid_ch': 12,
    'out_ch': 3,
}
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 RSU5 — 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 RSU5 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
RSU6 Ran
ran on a synthesized input · constructed an object; no result was computed
xuebinqin/DIS/IS-Net/models/isnet.py
code served (permissive licence) · get_code("dec5dbb951aa5aa8")
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, RSU6: 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)

ARGS = {
    'in_ch': 3,
    'mid_ch': 12,
    'out_ch': 3,
}
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 RSU6 — 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 RSU6 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
RSU7 Ran
ran on a synthesized input · constructed an object; no result was computed
xuebinqin/DIS/IS-Net/models/isnet.py
code served (permissive licence) · get_code("5f39ebc09fa09f33")
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, RSU7: 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)

ARGS = {
    'in_ch': 3,
    'mid_ch': 12,
    'out_ch': 3,
    'img_size': 512,
}
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 RSU7 — 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 RSU7 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
_upsample_like Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
xuebinqin/DIS/IS-Net/models/isnet.py
code served (permissive licence) · get_code("bfdbf21066b9aaf1")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 4, dtype family float, shape (2, 4, 6, 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
import torch.nn.functional as F

torch.manual_seed(0)

# The function _upsample_like takes two tensor arguments:
# - src: source tensor to upsample
# - tar: target tensor whose spatial dimensions (shape[2:]) define the target size

# Create a source tensor with shape (batch=2, channels=4, height=3, width=5)
src = torch.randn(2, 4, 3, 5)

# Create a target tensor with shape (batch=2, channels=4, height=6, width=8)
# The function will upsample src to match tar's spatial dimensions (6, 8)
tar = torch.randn(2, 4, 6, 8)

ARGS = {
    'src': src,
    'tar': tar
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 4, dtype family float, shape (2, 4, 6, 8) — shape (2, 4, 6, 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: none by the harness; any seed is in the setup block
muti_loss_fusion Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
xuebinqin/DIS/IS-Net/models/isnet.py
code served (permissive licence) · get_code("2ca2e4925d3db933")
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, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

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

torch.manual_seed(0)

# Create preds: list of tensors with different spatial dimensions
# Each pred should be shape (batch, channels, height, width)
# We'll make them progressively smaller to test the interpolation logic
pred0 = torch.sigmoid(torch.randn(2, 1, 8, 8))  # batch=2, channels=1, h=8, w=8
pred1 = torch.sigmoid(torch.randn(2, 1, 4, 4))  # batch=2, channels=1, h=4, w=4
pred2 = torch.sigmoid(torch.randn(2, 1, 2, 2))  # batch=2, channels=1, h=2, w=2

preds = [pred0, pred1, pred2]

# Create target: shape should match the first pred's spatial dimensions
# for at least one case, but we make it match pred0
target = torch.randint(0, 2, (2, 1, 8, 8)).float()  # batch=2, channels=1, h=8, w=8

ARGS = {
    "preds": preds,
    "target": target,
}
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, ndim 1, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
muti_loss_fusion_kl Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
xuebinqin/DIS/IS-Net/models/isnet.py
code served (permissive licence) · get_code("f8c4571c67857327")
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, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

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

torch.manual_seed(0)

# Create prediction tensors - list of tensors with different spatial sizes
preds = [
    torch.sigmoid(torch.randn(2, 1, 8, 8)),  # batch=2, channels=1, height=8, width=8
    torch.sigmoid(torch.randn(2, 1, 4, 4)),  # batch=2, channels=1, height=4, width=4
    torch.sigmoid(torch.randn(2, 1, 2, 2)),  # batch=2, channels=1, height=2, width=2
]

# Create target tensor
target = torch.sigmoid(torch.randn(2, 1, 8, 8))  # batch=2, channels=1, height=8, width=8

# Create deep feature tensors (dfs) - these will be compared with fs
dfs = [
    torch.randn(2, 4),  # batch=2, features=4
    torch.randn(2, 5),  # batch=2, features=5
]

# Create shallow feature tensors (fs) - same shapes as dfs
fs = [
    torch.randn(2, 4),  # batch=2, features=4
    torch.randn(2, 5),  # batch=2, features=5
]

# Mode parameter
mode = 'MSE'

ARGS = {
    'preds': preds,
    'target': target,
    'dfs': dfs,
    'fs': fs,
    'mode': mode,
}
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, ndim 1, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
ISNetDIS Not yet run
unverified (harvested, no recorded run)
xuebinqin/DIS/IS-Net/models/isnet.py
code served (permissive licence) · get_code("67b132694943823c")
How it ran

no invocation recorded.

Also present in other papers' code

Of 10 samples lifted from this paper's linked repositories, 10 carry a function identity and 0 do not. Of the 10 distinct functions with an identity, 5 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 5 later) and 5 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
RSU4
brick of this paper's assembly: model
1 other paper
1 later
2504.12739 (2025-04)
Mask Image Watermarking
RSU4F
brick of this paper's assembly: model
1 other paper
1 later
2504.12739 (2025-04)
Mask Image Watermarking
RSU5
brick of this paper's assembly: model
1 other paper
1 later
2504.12739 (2025-04)
Mask Image Watermarking
RSU6
brick of this paper's assembly: model
1 other paper
1 later
2504.12739 (2025-04)
Mask Image Watermarking
_upsample_like
brick of this paper's assembly: helper
1 other paper
1 later · 1 only as an assembly brick
2504.12739 (2025-04)
Mask Image Watermarking

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

We present a systematic study on a new task called dichotomous image segmentation (DIS) , which aims to segment highly accurate objects from natural images. To this end, we collected the first large-scale DIS dataset, called DIS5K, which contains 5,470 high-resolution (e.g., 2K, 4K or larger) images covering camouflaged, salient, or meticulous objects in various backgrounds. DIS is annotated with extremely finegrained labels. Besides, we introduce a simple intermediate supervision baseline (IS-Net) using both feature-level and mask-level guidance for DIS model training. IS-Net outperforms various cutting-edge baselines on the proposed DIS5K, making it a general self-learned supervision network that can facilitate future research in DIS. Further, we design a new metric called human correction efforts (HCE) which approximates the number of mouse clicking operations required to correct the false positives and false negatives. HCE is utilized to measure the gap between models and real-world applications and thus can complement existing metrics. Finally, we conduct the largest-scale benchmark, evaluating 16 representative segmentation models, providing a more insightful discussion regarding object complexities, and showing several potential applications (e.g., background removal, art design, 3D reconstruction). Hoping these efforts can open up promising directions for both academic and industries. Project page: https://xuebinqin.github.io/dis/index.html.

For agents

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

get_harvested_code_for_paper("2203.03041")
get_code_for_paper("2203.03041")
have("2203.03041")

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

curl https://syntology.ai/api/ran/2203.03041.json

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

[![Syntology run record](https://syntology.ai/api/ran/2203.03041.svg)](https://syntology.ai/paper/2203.03041)

Connect an agent — have() is free.