SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2303.11989 · ICCV · 2023

Text2Room: Extracting Textured 3D Meshes from 2D Text-to-Image Models

Andrew Owens, Justin Johnson, Matthias Nießner, Ang Cao, Lukas Höllein

arXiv · PDF · Open in the Atlas

Code that ran

13 of 13 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 13 have no contract verdict; 0 failures of our instrument; 0 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/2303.11989.json · badge: /api/ran/2303.11989.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
lukashoel/text2room canonical (classifier label)
classifier basis not recorded on the link
13 ran · 0 no recorded run
FunctionStatusWhere it lives
edge_threshold_filter Ran
ran on a synthesized input
lukashoel/text2room/model/mesh_fusion/render.py
code served (permissive licence) · get_code("1bf43bc41fcd3fa4")
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 int, shape (3, 0) 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)

# Create small test tensors
# vertices: (3, N) - 3D coordinates of N vertices
vertices = torch.randn(3, 8, dtype=torch.float32)

# faces: (3, M) - M triangular faces, each face has 3 vertex indices
# Create valid face indices pointing to vertices (0-7)
faces = torch.tensor([
    [0, 1, 2, 3],
    [1, 2, 3, 4],
    [2, 3, 4, 5]
], dtype=torch.long)

edge_threshold = 0.5

ARGS = {
    "vertices": vertices,
    "faces": faces,
    "edge_threshold": edge_threshold
}
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 int, shape (3, 0) — shape (3, 0) 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_pixel_grids Ran
ran on a synthesized input
lukashoel/text2room/model/mesh_fusion/util.py
code served (permissive licence) · get_code("721d622f67a06ca3")
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 (3, 20) 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 = {
    "height": 4,
    "width": 5,
    "reverse": False,
}
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 (3, 20) — shape (3, 20) 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
img_to_pts Ran
ran on a synthesized input
lukashoel/text2room/model/mesh_fusion/util.py
code served (permissive licence) · get_code("ca7fb024c005965d")
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 (3, 6) 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:
# img_to_pts(height, width, depth, K=torch.eye(3))
# - height: int, used for get_pixel_grids
# - width: int, used for get_pixel_grids
# - depth: tensor, needs .device attribute, reshaped to (1, -1), multiplied with pts
# - K: tensor, used with torch.inverse(K).mm(pts), so needs to be a square matrix
#   that matches the shape of pts (3, height*width)

# Construct arguments:
height = 2
width = 3
depth = torch.randn(1, height * width)
K = torch.eye(3)

ARGS = {
    "height": height,
    "width": width,
    "depth": depth,
    "K": K,
}
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 (3, 6) — shape (3, 6) 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
interpolate_noise Ran
ran on a synthesized input
lukashoel/text2room/model/trajectories/pose_noise_util.py
code served (permissive licence) · get_code("eeae9901967f7f80")
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 1, dtype family float, shape (20,) 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

# Construct arguments for interpolate_noise function
# The function expects:
# - n: a numpy array (used with n[-1], n[i], n[i+1], n.shape[0])
# - steps: an integer (used with num=steps parameter)

# Create a small numpy array with shape (4,) - 4 waypoints to interpolate between
n = np.array([0.0, 1.0, 2.0, 3.0])

# Number of interpolation steps between each pair of points
steps = 5

ARGS = {
    "n": n,
    "steps": steps
}
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 ndarray, ndim 1, dtype family float, shape (20,) — shape (20,) 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
parse_seen_poses Ran
ran on a synthesized input
lukashoel/text2room/model/trajectories/convert_to_nerf_convention.py
code served (permissive licence) · get_code("a1edc86df5512e7e")
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 dict. 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 numpy as np
import json
import tempfile
import os

torch.manual_seed(0)

# Create a temporary file with pose data in JSON format
temp_dir = tempfile.mkdtemp()
pose_file_path = os.path.join(temp_dir, "poses.json")

# Create sample pose data
pose_data = {
    "0": [[1.0, 0.0, 0.0, 0.5], [0.0, 1.0, 0.0, 0.5], [0.0, 0.0, 1.0, 0.5]],
    "1": [[0.9, 0.1, 0.0, 0.3], [0.0, 0.9, 0.1, 0.4], [0.0, 0.0, 1.0, 0.6]],
    "2": [[0.8, 0.2, 0.0, 0.2], [0.0, 0.8, 0.2, 0.3], [0.0, 0.0, 1.0, 0.7]]
}

with open(pose_file_path, 'w') as f:
    json.dump(pose_data, f)

ARGS = {
    "pose_file_path": pose_file_path
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
pil_to_torch Ran
ran on a synthesized input
lukashoel/text2room/model/utils/utils.py
code served (permissive licence) · get_code("23c6f8663a0b0977")
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 (3, 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
import numpy as np
from PIL import Image

torch.manual_seed(0)

# Create a small PIL Image (4x3 RGB)
img_array = np.random.randint(0, 256, size=(4, 3, 3), dtype=np.uint8)
pil_img = Image.fromarray(img_array, mode='RGB')

# Device set to CPU
device = 'cpu'

# Normalize flag
normalize = True

ARGS = {
    'img': pil_img,
    'device': device,
    'normalize': normalize
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, 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 3, dtype family float, shape (3, 4, 3) — shape (3, 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
pose_spherical Ran
ran on a synthesized input
lukashoel/text2room/model/trajectories/trajectory_util.py
code served (permissive licence) · get_code("87c4c2474bd1e51d")
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 (4, 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 numpy as np
import torch

torch.manual_seed(0)

def trans_t(t):
    return torch.Tensor([[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 1, t], [0, 0, 0, 1]]).float()

def rot_phi(phi):
    return torch.Tensor([[1, 0, 0, 0], [0, np.cos(phi), -np.sin(phi), 0], [0, np.sin(phi), np.cos(phi), 0], [0, 0, 0, 1]]).float()

def rot_theta(th):
    return torch.Tensor([[np.cos(th), 0, -np.sin(th), 0], [0, 1, 0, 0], [np.sin(th), 0, np.cos(th), 0], [0, 0, 0, 1]]).float()

def pose_spherical(theta, phi, radius):
    c2w = trans_t(radius)
    c2w = rot_phi(phi / 180.0 * np.pi) @ c2w
    c2w = rot_theta(theta / 180.0 * np.pi) @ c2w
    c2w[0:3, 3] = -1 * c2w[0:3, 3]
    return c2w

ARGS = {
    "theta": 45.0,
    "phi": 30.0,
    "radius": 4.0
}
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 2, dtype family float, shape (4, 4) — shape (4, 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
prepare_input Ran
ran on a synthesized input
lukashoel/text2room/model/iron_depth/predict_depth.py
code served (permissive licence) · get_code("40fe3e5b96d9ed64")
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 dict. 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 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 PIL import Image
from torchvision import transforms

torch.manual_seed(0)

# Create a small test image (RGB, 32x24 pixels)
test_image_array = np.random.randint(0, 256, (24, 32, 3), dtype=np.uint8)
test_image = Image.fromarray(test_image_array, mode='RGB')

# Create a valid intrinsic matrix K
K_matrix = np.array([
    [500.0, 0.0, 16.0],
    [0.0, 500.0, 12.0],
    [0.0, 0.0, 1.0]
], dtype=np.float32)

ARGS = {
    'image': test_image,
    'K': K_matrix,
    'device': 'cpu'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy, torch, torchvision — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
pts_to_image Ran
ran on a synthesized input
lukashoel/text2room/model/mesh_fusion/util.py
code served (permissive licence) · get_code("36674b1cdb0eab60")
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 3, 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)

# pts: tensor that will be multiplied by RT[:3, :3] and have RT[:3, 3:4] added
# RT[:3, :3] is 3x3, so pts must be 3xN
# K is 3x3 and multiplies wrld_X which is 3xN
pts = torch.randn(3, 2)

# K: camera intrinsic matrix, 3x3
# Used as K.mm(wrld_X) where wrld_X is 3x2
K = torch.randn(3, 3)

# RT: rotation-translation matrix
# Used as RT[:3, :3].mm(pts) + RT[:3, 3:4]
# RT[:3, :3] is 3x3, RT[:3, 3:4] is 3x1
# So RT should be at least 3x4
RT = torch.randn(3, 4)

ARGS = {
    "pts": pts,
    "K": K,
    "RT": RT,
}
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 3, 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
sample_noise Ran
ran on a synthesized input
lukashoel/text2room/model/trajectories/pose_noise_util.py
code served (permissive licence) · get_code("e6fffec018d7dd86")
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 3, 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 numpy as np

ARGS = {
    "n": 5,
    "r_max": 2.0,
    "t_max": 3.0,
}
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, ndim 3, 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
scale_shift_linear Ran
ran on a synthesized input
lukashoel/text2room/model/depth_alignment/depth_alignment.py
code served (permissive licence) · get_code("593f9e065a7464c0")
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 (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)

# Create small tensors with distinct dimensions
H, W = 4, 3

# rendered_depth: (H, W) tensor with positive values
rendered_depth = torch.rand(H, W) + 0.5  # Ensure positive values

# predicted_depth: (H, W) tensor with positive values
predicted_depth = torch.rand(H, W) + 0.5  # Ensure positive values

# mask: (H, W) binary tensor indicating valid points
mask = torch.randint(0, 2, (H, W)).bool()
# Ensure at least one valid point so mask.sum() > 0
mask[0, 0] = True

ARGS = {
    "rendered_depth": rendered_depth,
    "predicted_depth": predicted_depth,
    "mask": mask,
    "fuse": 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 2, dtype family float, shape (4, 3) — shape (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
to_degrees Ran
ran on a synthesized input
lukashoel/text2room/model/trajectories/pose_noise_util.py
code served (permissive licence) · get_code("580df8e07f77cad3")
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 1, dtype family float, shape (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 numpy as np

def to_degrees(x):
    return x * 180.0 / np.pi

# Create a small numpy array representing angles in radians
x = np.array([0.0, np.pi/2, np.pi, 2*np.pi])

ARGS = {
    "x": x
}
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 ndarray, ndim 1, dtype family float, shape (4,) — shape (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
visualize_depth_numpy Ran
ran on a synthesized input
lukashoel/text2room/model/utils/utils.py
code served (permissive licence) · get_code("a9b9a982a85106e3")
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 cv2
import numpy as np

# Create a small depth map (H=4, W=3)
depth = np.random.rand(4, 3) * 10.0

# Test with None minmax (will auto-compute)
ARGS = {
    "depth": depth,
    "minmax": None,
    "cmap": cv2.COLORMAP_TURBO,
}
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 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

Also present in other papers' code

Of 13 samples lifted from this paper's linked repositories, 13 carry a function identity and 0 do not. Of the 13 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 13 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

"a living room with lots of bookshelves, couches, and small tables" Iterative Generation (a) 3D Mesh Generation from Text (b) Rendered Image + Mesh Figure 1. Textured 3D mesh generation from text prompts. We generate textured 3D meshes from a given text prompt using 2D text-to-image models. (a) The scene is iteratively created from different viewpoints (marked in blue). (b) Our generated mesh contains compelling textures and geometry. We remove the ceiling in the top-down views for better visualization of the scene layout.

For agents

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

get_harvested_code_for_paper("2303.11989")
get_code_for_paper("2303.11989")
have("2303.11989")

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

curl https://syntology.ai/api/ran/2303.11989.json

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

[![Syntology run record](https://syntology.ai/api/ran/2303.11989.svg)](https://syntology.ai/paper/2303.11989)

Connect an agent — have() is free.