SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2104.06394 · 2021

All you need are a few pixels: semantic segmentation with PixelPick

arXiv · PDF · Open in the Atlas

Code that ran

9 of 11 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 7 have no contract verdict; 2 failures of our instrument; 2 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/2104.06394.json · badge: /api/ran/2104.06394.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
NoelShin/PixelPick canonical (classifier label)
classifier basis: census
4 ran · 1 no recorded run
yazdaniamir38/madbal community (classifier label)
classifier basis not recorded on the link
5 ran · 1 no recorded run
FunctionStatusWhere it lives
build_aspp Ran
ran; our contract draft was wrong, not the code
NoelShin/PixelPick/networks/aspp.py
code served (permissive licence) · get_code("c704648bd87c532f")
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 ASPP. 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 = {
    'backbone': 'resnet',
    'output_stride': 16,
    '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 ASPP — a ASPP output has no array shape of its own; none 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
conv_bn Ran
ran on a synthesized input
NoelShin/PixelPick/networks/mobilenet_v2.py
code served (permissive licence) · get_code("8cf426bba4e98da5")
How it ran

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

import torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    'inp': 3,
    'oup': 16,
    'stride': 1,
    '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 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
fixed_padding Ran
ran on a synthesized input
NoelShin/PixelPick/networks/mobilenet_v2.py
code served (permissive licence) · get_code("b34c12741c44b60c")
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, 7, 9) 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)

# Construct a small 4D tensor (batch, channels, height, width)
inputs = torch.randn(2, 4, 3, 5)

# Construct scalar parameters
kernel_size = 3
dilation = 2

ARGS = {
    "inputs": inputs,
    "kernel_size": kernel_size,
    "dilation": dilation,
}
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, 7, 9) — shape (2, 4, 7, 9) 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
gather_previous_query_files Ran
ran; our contract draft was wrong, not the code
NoelShin/PixelPick/query.py
code served (permissive licence) · get_code("693791a46cc3f42c")
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 tempfile
from pathlib import Path
from typing import List, Dict

# Create a temporary directory structure with test query files
temp_dir = tempfile.mkdtemp()
base_path = Path(temp_dir)

# Create nested directory structure with queries.pkl files
(base_path / "subdir1").mkdir(exist_ok=True)
(base_path / "subdir1" / "queries.pkl").touch()

(base_path / "subdir2" / "nested").mkdir(parents=True, exist_ok=True)
(base_path / "subdir2" / "nested" / "queries.pkl").touch()

(base_path / "queries.pkl").touch()

# Also create a queries.json file to test ext parameter
(base_path / "subdir1" / "queries.json").touch()

ARGS = {
    "dir_base": temp_dir,
    "ext": "pkl"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type 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
get_weights Ran
ran on a synthesized input
yazdaniamir38/madbal/utils/losses.py
code served (permissive licence) · get_code("c204954f25139fdd")
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 1, dtype family float, shape (3,) as a later run recorded it. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

torch.manual_seed(0)

# Create a target tensor with class labels and some 255 values (ignore index)
# Shape: (8,) - small batch
target = torch.tensor([0, 1, 2, 0, 1, 255, 2, 0], dtype=torch.long)

n_classes = 3

ARGS = {
    "target": target,
    "n_classes": n_classes,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 1, dtype family float, shape (3,) — shape (3,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
make_one_hot Ran
ran on a synthesized input
yazdaniamir38/madbal/utils/losses.py
code served (permissive licence) · get_code("3d38fe989e7b40c1")
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, 5, 4, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# labels: a tensor of shape (batch_size, 1, height, width) with class indices
# The function calls: labels.size()[0], labels.size()[2], labels.size()[3], labels.data, labels.device
batch_size = 2
height = 4
width = 3
num_classes = 5

labels = torch.randint(0, num_classes, (batch_size, 1, height, width))

ARGS = {
    "labels": labels,
    "classes": num_classes
}
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, 5, 4, 3) — shape (2, 5, 4, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
mse_loss Ran
ran on a synthesized input
yazdaniamir38/madbal/utils/losses.py
code served (permissive licence) · get_code("b68a8578f8b27ceb")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# Create small tensors for input and target
# input: predictions (batch_size=2, features=4)
# target: ground truth labels (batch_size=2, features=4)
input_tensor = torch.randn(2, 4)
target_tensor = torch.randn(2, 4)

# ignored_index: a value that marks elements to ignore in the loss calculation
ignored_index = -1

# reduction: how to aggregate the loss ('mean' or 'None')
reduction = 'mean'

ARGS = {
    'input': input_tensor,
    'target': target_tensor,
    'ignored_index': ignored_index,
    'reduction': reduction
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
multi_scale_predict Ran
ran on a synthesized input
yazdaniamir38/madbal/inference_VOC.py
code served (permissive licence) · get_code("d9286f5f6bd943b4")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 3, dtype family float, shape (5, 8, 12) 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
import numpy as np

torch.manual_seed(0)

# Create a minimal model that takes an input and produces a segmentation output
class MinimalSegmentationModel(nn.Module):
    def __init__(self, num_classes):
        super().__init__()
        self.num_classes = num_classes
        self.conv1 = nn.Conv2d(3, 16, kernel_size=3, padding=1)
        self.conv2 = nn.Conv2d(16, num_classes, kernel_size=3, padding=1)
    
    def forward(self, x):
        x = self.conv1(x)
        x = self.conv2(x)
        return x

# Create model instance
num_classes = 5
model = MinimalSegmentationModel(num_classes=num_classes)
model.eval()

# Create a small input image tensor (batch_size=1, channels=3, height=8, width=12)
image = torch.randn(1, 3, 8, 12)

# Define scales for multi-scale prediction
scales = [0.75, 1.0, 1.25]

# Device is CPU
device = 'cpu'

# Flip parameter
flip = True

ARGS = {
    'model': model,
    'image': image,
    'scales': scales,
    'num_classes': num_classes,
    'device': device,
    'flip': flip
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family float, shape (5, 8, 12) — shape (5, 8, 12) 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
pad_image Ran
ran on a synthesized input
yazdaniamir38/madbal/inference_VOC.py
code served (permissive licence) · get_code("b62891686b976239")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 3, dtype family float, shape (2, 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)

# Create a small image tensor: batch_size=2, channels=3, height=4, width=5
img = torch.randn(2, 4, 5)

# Define target size as a tuple (height, width)
target_size = (6, 8)

ARGS = {
    "img": img,
    "target_size": target_size,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 6, 8) — shape (2, 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
merge_previous_query_files Not yet run
unverified (harvested, no recorded run)
NoelShin/PixelPick/query.py
code served (permissive licence) · get_code("116373e2bac7614d")
How it ran

no invocation recorded.

sliding_predict Not yet run
unverified (harvested, no recorded run)
yazdaniamir38/madbal/inference_VOC.py
code served (permissive licence) · get_code("ae60cffe1b0b7d5b")
How it ran

no invocation recorded.

Also present in other papers' code

Of 11 samples lifted from this paper's linked repositories, 11 carry a function identity and 0 do not. Of the 11 distinct functions with an identity, 3 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 0 the same month, 3 later) and 8 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
build_aspp 6 other papers
3 dated earlier, 3 later
1802.02611 (2018-02)
Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation
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
and 3 more not named here
fixed_padding 3 other papers
2 dated earlier, 1 later
1901.04056 (2019-01)
The Liver Tumor Segmentation Benchmark (LiTS)
2003.05970 (2020-03)
LiDAR guided Small obstacle Segmentation
2109.01801 (2021-09)
Dual Transfer Learning for Event-based End-task Prediction via Pluggable Event to Image Translation
conv_bn 2 other papers
1 dated earlier, 1 later
2003.05970 (2020-03)
LiDAR guided Small obstacle Segmentation
2109.01801 (2021-09)
Dual Transfer Learning for Event-based End-task Prediction via Pluggable Event to Image Translation

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

A central challenge for the task of semantic segmentation is the prohibitive cost of obtaining dense pixel-level annotations to supervise model training. In this work, we show that in order to achieve a good level of segmentation performance, all you need are a few well-chosen pixel labels. We make the following contributions: (i) We investigate the novel semantic segmentation setting in which labels are supplied only at sparse pixel locations, and show that deep neural networks can use a handful of such labels to good effect; (ii) We demonstrate how to exploit this phenomena within an active learning framework, termed PixelPick, to radically reduce labelling cost, and propose an efficient "mouse-free" annotation strategy to implement our approach; (iii) We conduct extensive experiments to study the influence of annotation diversity under a fixed budget, model pretraining, model capacity and the sampling mechanism for picking pixels in this low annotation regime; (iv) We provide comparisons to the existing state of the art in semantic segmentation with active learning, and demonstrate comparable performance with up to two orders of magnitude fewer pixel annotations on the CamVid, Cityscapes and PASCAL VOC 2012 benchmarks; (v) Finally, we evaluate the efficiency of our annotation pipeline and its sensitivity to annotator error to demonstrate its practicality.

For agents

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

get_harvested_code_for_paper("2104.06394")
get_code_for_paper("2104.06394")
have("2104.06394")

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

curl https://syntology.ai/api/ran/2104.06394.json

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

[![Syntology run record](https://syntology.ai/api/ran/2104.06394.svg)](https://syntology.ai/paper/2104.06394)

Connect an agent — have() is free.