SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2202.00185 · 2022

LayoutEnhancer: Generating Good Indoor Layouts from Imperfect Data

arXiv · PDF · Open in the Atlas

Code that ran

5 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 5 have no contract verdict; 0 failures of our instrument; 3 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/2202.00185.json · badge: /api/ran/2202.00185.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
kleimertu/humancentriclayouts canonical (classifier label)
classifier basis not recorded on the link
5 ran · 3 no recorded run
FunctionStatusWhere it lives
center_furniture Ran
ran on a synthesized input
kleimertu/humancentriclayouts/src/main_functions.py
code served (permissive licence) · get_code("bb6470b953dda7e8")
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

torch.manual_seed(0)

class Furniture:
    def __init__(self, pos, dim, coarse_category, front_vec):
        self.pos = pos
        self.dim = dim
        self.coarse_category = coarse_category
        self._front_vec = front_vec
    
    def world_front(self):
        return self._front_vec

furniture_list = [
    Furniture(
        pos=torch.tensor([1.0, 2.0]),
        dim=torch.tensor([3.0, 4.0]),
        coarse_category='table',
        front_vec=torch.tensor([0.8, 0.6])
    ),
    Furniture(
        pos=torch.tensor([-1.0, 0.5]),
        dim=torch.tensor([2.0, 2.5]),
        coarse_category='window',
        front_vec=torch.tensor([1.0, 0.0])
    ),
]

ARGS = {
    'furniture': furniture_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 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
filter_furniture Ran
ran on a synthesized input
kleimertu/humancentriclayouts/src/main_functions.py
code served (permissive licence) · get_code("2c5d0de3135d0d08")
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

torch.manual_seed(0)

class Furniture:
    def __init__(self, coarse_category, pos, dim, front):
        self.coarse_category = coarse_category
        self.pos = pos
        self.dim = dim
        self._front = front
    
    def world_front(self):
        return self._front

# Create a small set of furniture objects
furniture_list = [
    Furniture(
        coarse_category='table',
        pos=torch.tensor([1.0, 2.0]),
        dim=torch.tensor([3.0, 4.0]),
        front=torch.tensor([0.0, 1.0])
    ),
    Furniture(
        coarse_category='window',
        pos=torch.tensor([5.0, 6.0]),
        dim=torch.tensor([2.0, 1.0]),
        front=torch.tensor([1.0, 0.0])
    ),
    Furniture(
        coarse_category='chair',
        pos=torch.tensor([-1.0, -2.0]),
        dim=torch.tensor([1.5, 2.5]),
        front=torch.tensor([0.707, 0.707])
    ),
    Furniture(
        coarse_category='door',
        pos=torch.tensor([0.0, 0.0]),
        dim=torch.tensor([1.0, 2.0]),
        front=torch.tensor([1.0, 0.0])
    ),
]

cat_valid = ['table', 'chair', 'window', 'door']

ARGS = {
    'furniture': furniture_list,
    'cat_valid': cat_valid,
    'centering': True
}
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
gaussian1d Ran
ran on a synthesized input
kleimertu/humancentriclayouts/src/main_functions.py
code served (permissive licence) · get_code("f117eee8d504b8c7")
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 (1, 10) 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)

ARGS = {
    'mu': 2.5,
    'sigma': 1.0,
    'res': 10,
    'device': 'cpu'
}
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 (1, 10) — shape (1, 10) 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
glare_par Ran
ran on a synthesized input
kleimertu/humancentriclayouts/src/evaluation.py
code served (permissive licence) · get_code("a8ed42b79afa84a1")
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, 4, 5) 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)

# Analyze the function:
# - pos: used in unsqueeze operations and subtraction with grid/light
# - light: used in unsqueeze operations and subtraction with pos
# - grid: used in unsqueeze operations and subtraction with pos
# - all_pairs: boolean flag controlling branching logic
#
# The function performs vector operations:
# - v = grid - pos (with unsqueezes for broadcasting)
# - lp = light - pos (with unsqueezes for broadcasting)
# - Both are normalized and dot products computed
# - Results passed through softmax and summed
#
# Shapes need to work with the unsqueeze/broadcasting logic:
# If all_pairs=True:
#   grid.shape = (batch, ..., 3) where ... represents spatial dims
#   pos.shape = (batch, ..., 3)
#   light.shape = (batch, ..., 3)
#   unsqueeze(grid, 1) and unsqueeze(pos, 2) for broadcasting

# Let's use:
# pos: (2, 4, 3) - batch=2, spatial=4, coords=3
# light: (2, 4, 3) - same structure
# grid: (2, 5, 3) - batch=2, different spatial size=5, coords=3
# This way dimensions are distinct and broadcasting works

pos = torch.randn(2, 4, 3)
light = torch.randn(2, 4, 3)
grid = torch.randn(2, 5, 3)

ARGS = {
    "pos": pos,
    "light": light,
    "grid": grid,
    "all_pairs": True
}
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, 4, 5) — shape (2, 4, 5) 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
lighting_par Ran
ran on a synthesized input
kleimertu/humancentriclayouts/src/evaluation.py
code served (permissive licence) · get_code("a7720d87742e78cf")
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, 3, 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

torch.manual_seed(0)

# Based on the function body:
# - pos: used in unsqueeze/subtract operations, appears to be shape (batch, 3) or (batch, features)
# - light: used in unsqueeze/subtract operations, similar to pos
# - grid: used in unsqueeze/subtract operations, appears to be (batch, num_points, 3) or similar
# - all_pairs: boolean flag

# Looking at the operations:
# torch.unsqueeze(grid, 1) - torch.unsqueeze(pos, 2)
# This suggests: grid is (batch, N, features), pos is (batch, features)
# Result should be broadcastable

# For all_pairs=True:
# v = torch.unsqueeze(grid, 1) - torch.unsqueeze(pos, 2)
# grid: (batch, N, features) -> unsqueeze(1) -> (batch, 1, N, features)
# pos: (batch, features) -> unsqueeze(2) -> (batch, features, 1)
# This doesn't match. Let me reconsider...

# Actually looking more carefully:
# unsqueeze(grid, 1): if grid is (B, N, 3), result is (B, 1, N, 3)
# unsqueeze(pos, 2): if pos is (B, M, 3), result is (B, M, 1, 3)
# Subtraction broadcasts to (B, M, N, 3)

# Let's use:
# batch_size = 2
# num_positions = 3
# num_grid_points = 4
# feature_dim = 3

pos = torch.randn(2, 3, 3)  # (batch, num_positions, features)
light = torch.randn(2, 3, 3)  # (batch, num_lights, features)
grid = torch.randn(2, 4, 3)  # (batch, num_grid_points, features)

ARGS = {
    "pos": pos,
    "light": light,
    "grid": grid,
    "all_pairs": True
}
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, 3, 4) — shape (2, 3, 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
train Not yet run
unverified (harvested, no recorded run)
kleimertu/humancentriclayouts/src/network.py
code served (permissive licence) · get_code("f19bd55602778439")
How it ran

no invocation recorded.

validate Not yet run
unverified (harvested, no recorded run)
kleimertu/humancentriclayouts/src/network.py
code served (permissive licence) · get_code("8118edf915e71390")
How it ran

no invocation recorded.

visibility_par Not yet run
unverified (harvested, no recorded run)
kleimertu/humancentriclayouts/src/evaluation.py
code served (permissive licence) · get_code("f1e67440c109a5df")
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, 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 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.

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 address the problem of indoor layout synthesis, which is a topic of continuing research interest in computer graphics. The newest works made significant progress using data-driven generative methods; however, these approaches rely on suitable datasets. In practice, desirable layout properties may not exist in a dataset, for instance, specific expert knowledge can be missing in the data. We propose a method that combines expert knowledge, for example, knowledge about ergonomics, with a data-driven generator based on the popular Transformer architecture. The knowledge is given as differentiable scalar functions, which can be used both as weights or as additional terms in the loss function. Using this knowledge, the synthesized layouts can be biased to exhibit desirable properties, even if these properties are not present in the dataset. Our approach can also alleviate problems of lack of data and imperfections in the data. Our work aims to improve generative machine learning for modeling and provide novel tools for designers and amateurs for the problem of interior layout creation.

For agents

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

get_harvested_code_for_paper("2202.00185")
get_code_for_paper("2202.00185")
have("2202.00185")

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

curl https://syntology.ai/api/ran/2202.00185.json

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

[![Syntology run record](https://syntology.ai/api/ran/2202.00185.svg)](https://syntology.ai/paper/2202.00185)

Connect an agent — have() is free.