SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1908.00821 · 2019

Learning Lightweight Lane Detection CNNs by Self Attention Distillation

arXiv · PDF · Open in the Atlas

Code that ran

7 of 8 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 4 have no contract verdict; 3 failures of our instrument; 1 no recorded run), measured 2026-11-21; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/1908.00821.json · badge: /api/ran/1908.00821.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
cardwing/Codes-for-Lane-Detection canonical (classifier label)
classifier basis: classifier_2025
1 ran · 1 no recorded run
NOHYC/autonomous_driving_car_project community (classifier label)
classifier basis not recorded on the link
6 ran · 0 no recorded run
FunctionStatusWhere it lives
assignment Ran
ran on a synthesized input
NOHYC/autonomous_driving_car_project/SIM/cvbridge_build_ws/devel/.private/project/_setup_util.py
code served (permissive licence) · get_code("6da2d91a099c5bee")
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 str. 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 platform

system = platform.system()

IS_WINDOWS = system == 'Windows'

def assignment(key, value):
    if not IS_WINDOWS:
        return 'export %s="%s"' % (key, value)
    else:
        return 'set %s=%s' % (key, value)


ARGS = {
    'key': 'MY_VAR',
    'value': 'my_value'
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
fanin_init Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
NOHYC/autonomous_driving_car_project/SIM/catkin_ws/src/project/src/ddpg_stage.py
pointer only (licence: MIT); the stored copy get_code serves, from fshamshirdar/pytorch-rdpg, is pointer only (licence: NONE) · get_code("0d0842514b1844a1")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from fshamshirdar/pytorch-rdpg, and GitHub detected no licence in that repository 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.

flip Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
cardwing/Codes-for-Lane-Detection/ERFNet-CULane-PyTorch/train_erfnet.py
code served (permissive licence) · get_code("1e6e2a7577f5c7c6")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned Tensor, ndim 3, dtype family float, shape (2, 4, 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 torch.nn as nn

torch.manual_seed(0)

# The function 'flip' takes two parameters: x (a tensor) and dim (an integer)
# It flips the tensor along the specified dimension
# x needs to be a tensor with at least 2 dimensions based on how it's used
# dim is an integer specifying which dimension to flip

x = torch.randn(2, 4, 3)  # Small tensor: batch=2, dim1=4, dim2=3
dim = 1  # Flip along dimension 1

ARGS = {
    "x": x,
    "dim": dim
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 4, 3) — shape (2, 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
get_cameramat_dist Ran
ran on a synthesized input
NOHYC/autonomous_driving_car_project/NX/catkin_ws/src/enet_ros/src/pub_camera.py
code served (permissive licence) · get_code("1784d626db4641c8")
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 pickle
import tempfile
import os
import numpy as np

# Create a temporary file with pickled camera calibration data
temp_dir = tempfile.gettempdir()
filename = os.path.join(temp_dir, "test_camera_calib.pkl")

# Create sample camera matrix, distortion coefficients, rvecs, and tvecs
mat = np.array([[500.0, 0.0, 320.0],
                [0.0, 500.0, 240.0],
                [0.0, 0.0, 1.0]], dtype=np.float32)

dist = np.array([[-0.1, 0.05, 0.001, 0.002]], dtype=np.float32)

rvecs = [np.array([0.1, 0.2, 0.3], dtype=np.float32),
         np.array([0.15, 0.25, 0.35], dtype=np.float32)]

tvecs = [np.array([1.0, 2.0, 3.0], dtype=np.float32),
         np.array([1.5, 2.5, 3.5], dtype=np.float32)]

# Pickle the data
with open(filename, 'wb') as f:
    pickle.dump((mat, dist, rvecs, tvecs), f)

ARGS = {
    "filename": filename
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — 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
mish Ran
ran; our contract draft was wrong, not the code
NOHYC/autonomous_driving_car_project/SIM/catkin_ws/src/project/src/ddpg_stage.py
code served (permissive licence) · get_code("55a0b1ed4884f9c2")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn.functional as F

torch.manual_seed(0)

# The mish function takes a single parameter x which is a tensor
# It applies torch.tanh and F.softplus operations on it
# So x should be a tensor that supports these operations

x = torch.randn(2, 4)

ARGS = {
    "x": x
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
prepend_env_variables Ran
ran on a synthesized input
NOHYC/autonomous_driving_car_project/SIM/cvbridge_build_ws/devel/.private/project/_setup_util.py
code served (permissive licence) · get_code("614a55e26da69252")
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 os
import tempfile

# Create temporary directories to simulate workspaces
temp_dir1 = tempfile.mkdtemp()
temp_dir2 = tempfile.mkdtemp()

# Create subdirectories for env_var_subfolders
os.makedirs(os.path.join(temp_dir1, 'lib'), exist_ok=True)
os.makedirs(os.path.join(temp_dir2, 'bin'), exist_ok=True)

# Set up the environment dictionary
environ = {
    'PATH': '/usr/bin:/bin',
    'LD_LIBRARY_PATH': '/usr/lib',
}

# Set up env_var_subfolders mapping variable names to subfolders
env_var_subfolders = {
    'CMAKE_PREFIX_PATH': '',
    'LD_LIBRARY_PATH': 'lib',
    'PATH': 'bin',
}

# Set up workspaces as a path separator-delimited string
workspaces = os.pathsep.join([temp_dir1, temp_dir2])

ARGS = {
    'environ': environ,
    'env_var_subfolders': env_var_subfolders,
    'workspaces': workspaces,
}
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
rollback_env_variables Ran
ran on a synthesized input
NOHYC/autonomous_driving_car_project/SIM/cvbridge_build_ws/devel/.private/project/_setup_util.py
code served (permissive licence) · get_code("8cd63ad2dbce2c60")
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 os
import copy

# Set up environment for testing
test_environ = {
    'CMAKE_PREFIX_PATH': '/workspace1:/workspace2:/opt/ros/fuerte/share',
    'PATH': '/usr/bin:/usr/local/bin',
    'PYTHONPATH': '/workspace1/lib:/workspace2/lib:/opt/ros/fuerte/lib',
    'LD_LIBRARY_PATH': '/workspace1/lib:/workspace2/lib'
}

# Create minimal workspace marker files in temp locations to simulate catkin workspaces
# Since we can't do real file I/O, we'll construct a test case where the marker files
# would exist but use include_non_existing=True in _get_workspaces

env_var_subfolders = {
    'PATH': ['bin'],
    'PYTHONPATH': ['lib'],
    'LD_LIBRARY_PATH': 'lib'
}

ARGS = {
    'environ': test_environ,
    'env_var_subfolders': env_var_subfolders
}
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
validate Not yet run
unverified (harvested, no recorded run)
cardwing/Codes-for-Lane-Detection/ERFNet-CULane-PyTorch/train_erfnet.py
code served (permissive licence) · get_code("0eda96b389501673")
How it ran

no invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 distinct functions with an identity, 2 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 2 later) and 6 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
mish 15 other papers
15 later
2001.07685 (2020-01)
FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence
2105.14148 (2021-05)
OpenMatch: Open-set Consistency Regularization for Semi-supervised Learning with Outliers
2110.03006 (2021-10)
Unsupervised Selective Labeling for More Effective Semi-Supervised Learning
and 12 more not named here
fanin_init 3 other papers
2 dated earlier, 1 later · 1 only as an assembly brick
1509.02971 (2015-09)
Continuous control with deep reinforcement learning
1805.07917 (2018-05)
Evolution-Guided Policy Gradient in Reinforcement Learning
2209.08646 (2022-09)
DeepTOP: Deep Threshold-Optimal Policy for MDPs and RMABs

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-21.

Repositories linked to this paper

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

Abstract

Training deep models for lane detection is challenging due to the very subtle and sparse supervisory signals inherent in lane annotations. Without learning from much richer context, these models often fail in challenging scenarios, e.g., severe occlusion, ambiguous lanes, and poor lighting conditions. In this paper, we present a novel knowledge distillation approach, i.e., Self Attention Distillation (SAD), which allows a model to learn from itself and gains substantial improvement without any additional supervision or labels. Specifically, we observe that attention maps extracted from a model trained to a reasonable level would encode rich contextual information. The valuable contextual information can be used as a form of 'free' supervision for further representation learning through performing topdown and layer-wise attention distillation within the network itself. SAD can be easily incorporated in any feedforward convolutional neural networks (CNN) and does not increase the inference time. We validate SAD on three popular lane detection benchmarks (TuSimple, CULane and BDD100K) using lightweight models such as ENet, ResNet-18 and ResNet-34. The lightest model, ENet-SAD, performs comparatively or even surpasses existing algorithms. Notably, ENet-SAD has 20 x fewer parameters and runs 10 x faster compared to the state-of-the-art SCNN, while still achieving compelling performance in all benchmarks. Our code is available at https://github.com/cardwing/Codes-for-Lane-Detection.

For agents

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

get_harvested_code_for_paper("1908.00821")
get_code_for_paper("1908.00821")
have("1908.00821")

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

curl https://syntology.ai/api/ran/1908.00821.json

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

[![Syntology run record](https://syntology.ai/api/ran/1908.00821.svg)](https://syntology.ai/paper/1908.00821)

Connect an agent — have() is free.