SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2203.03339 · 2022

L2CS-Net: Fine-Grained Gaze Estimation in Unconstrained Environments

arXiv · PDF · Open in the Atlas

Code that ran

8 of 10 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; 1 honoured the contract we drafted, 0 violated it and 3 have no contract verdict; 4 failures of our instrument; 2 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/2203.03339.json · badge: /api/ran/2203.03339.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
ahmednull/l2cs-net community (classifier label)
classifier basis not recorded on the link
5 ran · 0 no recorded run
copy not recorded community
not in the paper's declared repositories
2 ran · 0 no recorded run
yakhyo/gaze-estimation extension (classifier label)
classifier basis: classifier_2025
1 ran · 2 no recorded run
FunctionStatusWhere it lives
atoi Ran
ran, honoured the contract we drafted
ahmednull/l2cs-net/l2cs/utils.py
pointer only (licence: MIT); the stored copy get_code serves, from harshmaru7/DilatedConv, is pointer only (licence: NONE) · get_code("6dbf1b2a5b901370")
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 harshmaru7/DilatedConv, 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.

conv1x1 Ran
ran; our contract draft was wrong, not the code
this paper's copy was not recorded; identical code first harvested from yakhyo/head-pose-estimation
pointer only · get_code("d97ed054442099c9")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. 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.

conv3x3 Ran
ran; our contract draft was wrong, not the code
this paper's copy was not recorded; identical code first harvested from yakhyo/head-pose-estimation
pointer only · get_code("7f22d053fb2c79d3")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. 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.

draw_bbox Ran
ran on a synthesized input
ahmednull/l2cs-net/l2cs/vis.py
code served (permissive licence) · get_code("1db8413dd5010d6d")
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 int, shape (480, 640, 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 cv2
import numpy as np

# Create a sample frame (image) - 480x640 RGB image
frame = np.zeros((480, 640, 3), dtype=np.uint8)

# Create a sample bounding box [x_min, y_min, x_max, y_max]
bbox = np.array([50.5, 75.3, 200.8, 300.2], dtype=np.float32)

ARGS = {
    "frame": frame,
    "bbox": bbox
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy — 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 int, shape (480, 640, 3) — shape (480, 640, 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
draw_gaze Ran
ran on a synthesized input
ahmednull/l2cs-net/l2cs/vis.py
code served (permissive licence) · get_code("c245dc2db42c51e1")
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 int, shape (480, 640, 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 cv2
import numpy as np

# Create a sample image (BGR format, typical for cv2)
image_in = np.random.randint(0, 256, (480, 640, 3), dtype=np.uint8)

# Parameters for the function
a = 100  # x coordinate of eye region
b = 150  # y coordinate of eye region
c = 50   # width/length parameter
d = 60   # height parameter

# Gaze angles (pitch, yaw) in radians
pitchyaw = np.array([0.3, -0.2])

# Optional parameters
thickness = 2
color = (255, 255, 0)  # BGR yellow
sclae = 2.0

ARGS = {
    'a': a,
    'b': b,
    'c': c,
    'd': d,
    'image_in': image_in,
    'pitchyaw': pitchyaw,
    'thickness': thickness,
    'color': color,
    'sclae': sclae
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy — 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 int, shape (480, 640, 3) — shape (480, 640, 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
natural_keys Ran
ran; our contract draft was wrong, not the code
ahmednull/l2cs-net/l2cs/utils.py
code served (permissive licence) · get_code("7052cf0436949e37")
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 re

def atoi(text):
    return int(text) if text.isdigit() else text

def natural_keys(text):
    """
    alist.sort(key=natural_keys) sorts in human order
    http://nedbatchelder.com/blog/200712/human_sorting.html
    (See Toothy's implementation in the comments)
    """
    return [atoi(c) for c in re.split('(\\d+)', text)]

# The function takes a single parameter: text (a string)
# It uses string methods: isdigit(), re.split()
# It returns a list of mixed ints and strings

ARGS = {
    "text": "file123name456.txt"
}
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
prep_input_numpy Ran
ran on a synthesized input
ahmednull/l2cs-net/l2cs/utils.py
code served (permissive licence) · get_code("bf642780353748df")
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 (1, 3, 896, 448) 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 numpy as np
import torch
from torchvision import transforms

torch.manual_seed(0)

# Create a small test image as numpy array (3-channel, height, width)
# Using distinct dimensions: 448x224x3 to match the resize operation in transformations
img_single = np.random.randint(0, 256, (448, 224, 3), dtype=np.uint8)

# Create a batch of images
img_batch = np.random.randint(0, 256, (2, 448, 224, 3), dtype=np.uint8)

# The function accepts either a single image (3D) or batch (4D)
# We'll test with a single image
device = 'cpu'

ARGS = {
    'img': img_single,
    'device': device
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch, torchvision — 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 (1, 3, 896, 448) — shape (1, 3, 896, 448) 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
reparameterize_model Ran
ran; our contract draft was wrong, not the code
yakhyo/gaze-estimation/models/mobileone.py
code served (permissive licence) · get_code("4283b387669166bd")
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 SimpleMobileOne. 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 copy
import torch
import torch.nn as nn

torch.manual_seed(0)

# Create a minimal MobileOne-like model with reparameterizable modules
class ReparameterizableBlock(nn.Module):
    """A simple block that has a reparameterize method."""
    def __init__(self):
        super().__init__()
        self.conv = nn.Conv2d(3, 4, kernel_size=3, padding=1)
        self.bn = nn.BatchNorm2d(4)
        self.reparameterized = False
    
    def reparameterize(self):
        """Simulate reparameterization by fusing branches."""
        self.reparameterized = True
    
    def forward(self, x):
        x = self.conv(x)
        x = self.bn(x)
        return x

class SimpleMobileOne(nn.Module):
    """Minimal MobileOne model for testing."""
    def __init__(self):
        super().__init__()
        self.block1 = ReparameterizableBlock()
        self.block2 = ReparameterizableBlock()
        self.linear = nn.Linear(4, 2)
    
    def forward(self, x):
        x = self.block1(x)
        x = self.block2(x)
        x = x.mean(dim=[2, 3])  # Global average pooling
        x = self.linear(x)
        return x

# Create the model
model = SimpleMobileOne()

ARGS = {
    "model": model
}
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 SimpleMobileOne — a SimpleMobileOne output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a SimpleMobileOne 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
create_mobileone_model Not yet run
unverified (harvested, no recorded run)
yakhyo/gaze-estimation/models/mobileone.py
code served (permissive licence) · get_code("4bddc67420125c05")
How it ran

no invocation recorded.

mobileone_s0 Not yet run
unverified (harvested, no recorded run)
yakhyo/gaze-estimation/models/mobileone.py
code served (permissive licence) · get_code("4df4b0135ebb8b1f")
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 (4 with at least one paper dated earlier, 0 the same month, 3 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
atoi 13 other papers
3 dated earlier, 10 later
1511.07122 (2015-11)
Multi-Scale Context Aggregation by Dilated Convolutions
1908.03706 (2019-08)
Exploiting temporal consistency for real-time video depth estimation
2104.08773 (2021-04)
Cross-Task Generalization via Natural Language Crowdsourcing Instructions
and 10 more not named here
natural_keys 12 other papers
2 dated earlier, 10 later
1511.07122 (2015-11)
Multi-Scale Context Aggregation by Dilated Convolutions
1908.03706 (2019-08)
Exploiting temporal consistency for real-time video depth estimation
2208.13645 (2022-08)
Finding Near-Optimal Weight Independent Sets at Scale
and 9 more not named here
conv1x1 4 other papers
4 dated earlier
1512.03385 (2015-12)
Deep Residual Learning for Image Recognition
1710.00925 (2017-10)
Fine-Grained Head Pose Estimation Without Keypoints
1806.01531 (2018-06)
Deep Mixture of Experts via Shallow Embedding
and 1 more not named here
conv3x3 4 other papers
4 dated earlier
1512.03385 (2015-12)
Deep Residual Learning for Image Recognition
1710.00925 (2017-10)
Fine-Grained Head Pose Estimation Without Keypoints
1806.01531 (2018-06)
Deep Mixture of Experts via Shallow Embedding
and 1 more not named here
reparameterize_model 3 other papers
3 later
2206.04040 (2022-06)
MobileOne: An Improved One millisecond Mobile Backbone
2309.03452 (2023-09)
Multimodal Guidance Network for Missing-Modality Inference in Content Moderation
2406.05850 (2024-06)
Scaling Graph Convolutions for Mobile Vision

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

Human gaze is a crucial cue used in various applications such as human-robot interaction and virtual reality. Recently, convolution neural network (CNN) approaches have made notable progress in predicting gaze direction. However, estimating gaze in-the-wild is still a challenging problem due to the uniqueness of eye appearance, lightning conditions, and the diversity of head pose and gaze directions. In this paper, we propose a robust CNN-based model for predicting gaze in unconstrained settings. We propose to regress each gaze angle separately to improve the per-angel prediction accuracy, which will enhance the overall gaze performance. In addition, we use two identical losses, one for each angle, to improve network learning and increase its generalization. We evaluate our model with two popular datasets collected with unconstrained settings. Our proposed model achieves state-of-the-art accuracy of 3.92° and 10.41° on MPIIGaze and Gaze360 datasets, respectively. We make our code open source at https://github.com/Ahmednull/L2CS-Net.

For agents

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

get_harvested_code_for_paper("2203.03339")
get_code_for_paper("2203.03339")
have("2203.03339")

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

curl https://syntology.ai/api/ran/2203.03339.json

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

[![Syntology run record](https://syntology.ai/api/ran/2203.03339.svg)](https://syntology.ai/paper/2203.03339)

Connect an agent — have() is free.