SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2103.14644 · ICCV · 2021

Planar Surface Reconstruction from Sparse Views

Jin Linyi, Andrew Qian, David Owens, Fouhey

arXiv · PDF · Open in the Atlas

Code that ran

17 of 18 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; 3 honoured the contract we drafted, 1 violated it and 0 have no contract verdict; 13 failures of our instrument; 1 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/2103.14644.json · badge: /api/ran/2103.14644.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
jinlinyi/SparsePlanes canonical (classifier label)
classifier basis: census
17 ran · 1 no recorded run
FunctionStatusWhere it lives
angle_error_vec Ran
ran, honoured the contract we drafted
jinlinyi/SparsePlanes/sparsePlane/sparseplane/evaluation/mp3d_evaluation.py
code served (permissive licence) · get_code("3c23388604dafa28")
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 (2,) 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

torch_seed = 0

# angle_error_vec computes angular error between pairs of vectors
# It expects v1 and v2 to be 2D arrays with shape (batch_size, vector_dim)
# The function sums element-wise products along axis=1, so vectors should be normalized or unit vectors

batch_size = 2
vector_dim = 4

# Create two sets of random vectors
# For meaningful angle computation, vectors should ideally be unit vectors or normalized
v1 = np.random.randn(batch_size, vector_dim).astype(np.float32)
v1 = v1 / np.linalg.norm(v1, axis=1, keepdims=True)  # normalize

v2 = np.random.randn(batch_size, vector_dim).astype(np.float32)
v2 = v2 / np.linalg.norm(v2, axis=1, keepdims=True)  # normalize

ARGS = {
    "v1": v1,
    "v2": v2,
}
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 (2,) — shape (2,) 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
conv2d Ran
ran; our contract draft was wrong, not the code
jinlinyi/SparsePlanes/sparsePlane/sparseplane/modeling/camera_net/camera_head.py
code served (permissive licence) · get_code("6584ae4f2836a5ff")
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 = {
    "in_channels": 3,
    "out_channels": 16,
    "kernel_size": 3,
    "stride": 1,
    "padding": None,
}
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
deconv2d Ran
ran; our contract draft was wrong, not the code
jinlinyi/SparsePlanes/sparsePlane/sparseplane/modeling/camera_net/camera_head.py
code served (permissive licence) · get_code("8772d6bd1907c0d8")
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
from torch import nn

torch.manual_seed(0)

ARGS = {
    'scale_factor': 2,
    'mode': 'nearest',
    'in_channels': 256,
    'out_channels': 128,
    'kernel_size': 3,
    'stride': 1,
    'padding': 1
}
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
face2vert Ran
ran; our contract draft was wrong, not the code
jinlinyi/SparsePlanes/data_preparation/convert_mp3d_ply.py
code served (permissive licence) · get_code("b58cfd2354528e6e")
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

# Analyze function usage:
# - face_id: used as scalar assignment target, appears to be an integer or array index
# - face: used with .max() method, indexed with [:, i] for i in range(3)
#   so face must be a 2D array with shape (N, 3) where N is number of faces

# Create small test data
face_id = 1  # scalar face identifier
face = np.array([[0, 1, 2],
                 [1, 2, 3]], dtype=np.int32)  # 2 faces, 3 vertices each, max vertex id is 3

ARGS = {
    "face_id": face_id,
    "face": face
}
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
fitPlane Ran
ran, honoured the contract we drafted
jinlinyi/SparsePlanes/data_preparation/fit_plane_ply.py
code served (permissive licence) · get_code("af8d8bdd59e32da6")
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 (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 numpy as np

# fitPlane expects a numpy array (points) as input
# The function checks if points.shape[0] == points.shape[1] (square matrix case)
# or uses lstsq for non-square case
# It performs matrix operations: np.linalg.solve or np.linalg.lstsq
# with np.ones(points.shape[0]) as the right-hand side

# Create a non-square points array (overdetermined system)
# Using shape (5, 3) so it will go to the lstsq branch
points = np.array([
    [1.0, 2.0, 3.0],
    [4.0, 5.0, 6.0],
    [7.0, 8.0, 9.0],
    [10.0, 11.0, 12.0],
    [13.0, 14.0, 15.0]
], dtype=np.float64)

ARGS = {
    "points": points
}
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 (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
fix_JSON Ran
ran; our contract draft was wrong, not the code
jinlinyi/SparsePlanes/data_preparation/fit_plane_ply.py
code served (permissive licence) · get_code("ef49b8c62716105b")
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 json

# The function fix_JSON takes a json_message parameter (a string)
# and attempts to parse it as JSON, with a fallback that removes backslashes

# Test case 1: Valid JSON string
valid_json_str = '{"key": "value", "number": 42}'

# Test case 2: JSON with escaped backslashes that need fixing
invalid_json_str = '{"path": "C:\\\\Users\\\\test"}'

# We'll use the valid JSON as the main test case
ARGS = {
    "json_message": valid_json_str
}
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 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
get_K_inv_dot_xy_1 Ran
ran, honoured the contract we drafted
jinlinyi/SparsePlanes/data_preparation/render_plane.py
code served (permissive licence) · get_code("87f16c9e7b3ad9d2")
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 (3, 480, 640) 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

ARGS = {
    'h': 480,
    'w': 640,
}
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 3, dtype family float, shape (3, 480, 640) — shape (3, 480, 640) 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_depth_err Ran
ran; our contract draft was wrong, not the code
jinlinyi/SparsePlanes/sparsePlane/sparseplane/evaluation/mp3d_evaluation.py
code served (permissive licence) · get_code("559d8fc02b1cddcc")
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

torch.manual_seed(0)

# Create small tensors for pred_depth and gt_depth
# Shape: (2, 4) - batch size 2, 4 depth values
pred_depth = torch.randn(2, 4)
gt_depth = torch.rand(2, 4) * 0.1  # Values between 0 and 0.1 to have some > 0.0001

# device parameter - None by default
device = None

ARGS = {
    "pred_depth": pred_depth,
    "gt_depth": gt_depth,
    "device": device,
}
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
get_hfov Ran
ran, violated the contract we drafted
jinlinyi/SparsePlanes/data_preparation/utils.py
code served (permissive licence) · get_code("8f50175e76de5c7b")
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 float32, 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 numpy as np

# The function expects:
# - intrinsics: a numpy array that can be reshaped to have at least 1 element (accessed via [0])
# - width: an integer (default 1280)

# Create a small intrinsics array representing camera focal length
# Typically this would be a 3x3 or 2x2 intrinsic matrix, but the function only needs the first element
intrinsics = np.array([[500.0, 0.0, 320.0],
                       [0.0, 500.0, 240.0],
                       [0.0, 0.0, 1.0]], dtype=np.float32)

ARGS = {
    "intrinsics": intrinsics,
    "width": 1280
}
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 float32, 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
l1LossMask Ran
ran; our contract draft was wrong, not the code
jinlinyi/SparsePlanes/sparsePlane/sparseplane/evaluation/mp3d_evaluation.py
code served (permissive licence) · get_code("45d9c8eeef02238c")
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

torch.manual_seed(0)

# Create small tensors for the function
# pred: predicted values
# gt: ground truth values
# mask: binary mask to select which elements to include in loss

pred = torch.randn(2, 4)
gt = torch.randn(2, 4)
mask = torch.randint(0, 2, (2, 4), dtype=torch.float32)

ARGS = {
    "pred": pred,
    "gt": gt,
    "mask": mask
}
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
mergePlanesNew Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
jinlinyi/SparsePlanes/data_preparation/fit_plane_ply.py
code served (permissive licence) · get_code("485d6c3c648a4454")
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 numpy as np

# Create a small set of 3D points (N points in 3D space)
points = np.array([
    [0.0, 0.0, 0.0],
    [1.0, 0.0, 1.0],
    [0.0, 1.0, 1.0],
    [1.0, 1.0, 2.0],
    [2.0, 0.0, 2.0],
    [2.0, 1.0, 3.0],
], dtype=np.float64)

# Create plane coefficients (ax + by + cz = 1, so plane is [a, b, c])
planes = [
    np.array([0.5, 0.3, 0.2], dtype=np.float64),
    np.array([0.1, 0.4, 0.3], dtype=np.float64),
]

# Point indices for each plane (which points belong to which plane)
planePointIndices = [
    np.array([0, 1, 2], dtype=np.int64),
    np.array([3, 4, 5], dtype=np.int64),
]

# Segments (using sets of segment IDs)
planeSegments = [
    {0, 1},
    {2, 3},
]

# Segment neighbors (mapping segment ID to neighboring segment IDs)
segmentNeighbors = {
    0: [2],
    1: [3],
    2: [0],
    3: [1],
}

# Number of planes range (min, max)
numPlanes = (1, 5)

# Optional parameters with default values
planeDiffThreshold = 0.05
planeAngleThreshold = 30
inlierThreshold = 0.9
planeAreaThreshold = 10
orthogonalThreshold = np.cos(np.deg2rad(60))
parallelThreshold = np.cos(np.deg2rad(30))
debug = False

ARGS = {
    'points': points,
    'planes': planes,
    'planePointIndices': planePointIndices,
    'planeSegments': planeSegments,
    'segmentNeighbors': segmentNeighbors,
    'numPlanes': numPlanes,
    'planeDiffThreshold': planeDiffThreshold,
    'planeAngleThreshold': planeAngleThreshold,
    'inlierThreshold': inlierThreshold,
    'planeAreaThreshold': planeAreaThreshold,
    'orthogonalThreshold': orthogonalThreshold,
    'parallelThreshold': parallelThreshold,
    'debug': debug,
}
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 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
mp3d2habitat Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
jinlinyi/SparsePlanes/data_preparation/render_plane.py
code served (permissive licence) · get_code("a6798588f1400f85")
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 2, dtype family float, shape (2, 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 numpy as np

# The function mp3d2habitat takes a 'planes' parameter
# Looking at the function body:
# - planes is converted to np.array and transposed
# - It's multiplied by a rotation matrix (3x3)
# - The result is transposed back
# This means planes should be a 2D array-like with shape (N, 3)
# where N is the number of planes and 3 are the coordinates

# Create a small sample planes array with shape (2, 3)
# representing 2 planes with 3 coordinates each
planes = np.array([
    [1.0, 2.0, 3.0],
    [4.0, 5.0, 6.0]
])

ARGS = {
    "planes": planes
}
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 2, dtype family float, shape (2, 3) — shape (2, 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
pad_masks Ran
ran; our contract draft was wrong, not the code
jinlinyi/SparsePlanes/sparsePlane/sparseplane/layers/mask_ops.py
code served (permissive licence) · get_code("19a63e894fd02add")
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 torch

torch.manual_seed(0)

# Create a small batch of masks with shape (B, M, M)
# B=2 (batch size), M=4 (mask dimension)
masks = torch.randint(0, 2, (2, 4, 4), dtype=torch.float32)

# Padding amount
padding = 1

ARGS = {
    "masks": masks,
    "padding": padding,
}
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 — 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
paste_mask_in_image_old Ran
ran; our contract draft was wrong, not the code
jinlinyi/SparsePlanes/sparsePlane/sparseplane/layers/mask_ops.py
code served (permissive licence) · get_code("3f6042fe5727b417")
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 (100, 150) 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

torch.manual_seed(0)

# Create a small mask tensor (Hmask, Wmask) with values in [0, 1]
mask = torch.rand(8, 6, dtype=torch.float32)

# Create a box tensor (4,) with x0, y0, x1, y1 coordinates
box = torch.tensor([10, 20, 25, 35], dtype=torch.float32)

# Image dimensions
img_h = 100
img_w = 150

# Threshold for binarization
threshold = 0.5

ARGS = {
    "mask": mask,
    "box": box,
    "img_h": img_h,
    "img_w": img_w,
    "threshold": threshold,
}
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 2, dtype family int, shape (100, 150) — shape (100, 150) 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
paste_masks_in_image Ran
ran; our contract draft was wrong, not the code
jinlinyi/SparsePlanes/sparsePlane/sparseplane/layers/mask_ops.py
code served (permissive licence) · get_code("8254e47b641929b1")
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, 24, 32) 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

torch.manual_seed(0)

# Small mask predictions: (N, Hmask, Wmask)
# N=2 instances, 4x4 mask resolution
masks = torch.rand(2, 4, 4)

# Bounding boxes: (N, 4) in format [x0, y0, x1, y1]
# Boxes within image bounds (0-32 for width, 0-24 for height)
boxes = torch.tensor([
    [5.0, 3.0, 15.0, 18.0],
    [10.0, 5.0, 20.0, 19.0]
], dtype=torch.float32)

# Image shape: (height, width)
image_shape = (24, 32)

# Threshold for binary mask conversion
threshold = 0.5

# NMS flag
nms = False

ARGS = {
    "masks": masks,
    "boxes": boxes,
    "image_shape": image_shape,
    "threshold": threshold,
    "nms": nms
}
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 3, dtype family float, shape (2, 24, 32) — shape (2, 24, 32) 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
remove_small_plane Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
jinlinyi/SparsePlanes/data_preparation/render_plane.py
code served (permissive licence) · get_code("dd689ce211ee4516")
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 2, dtype family int, shape (5, 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 cv2
import numpy as np

# Create a small plane instance ID array
# The function expects a 2D array where different values represent different plane instances
plane_instance_id = np.array([
    [0, 0, 1, 1, 2, 2],
    [0, 0, 1, 1, 2, 2],
    [0, 0, 1, 1, 2, 2],
    [0, 0, 3, 3, 2, 2],
    [0, 0, 3, 3, 2, 2],
], dtype=np.int32)

ARGS = {
    'plane_instance_id': plane_instance_id,
    'bkgd': 0,
}
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 2, dtype family int, shape (5, 6) — shape (5, 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
sem_seg_postprocess Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
jinlinyi/SparsePlanes/sparsePlane/sparseplane/modeling/postprocessing.py
code served (permissive licence) · get_code("7d1dddea95f2a435")
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 (4, 24, 32) 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 semantic segmentation prediction logits tensor
# Shape: (C, H, W) where C is number of classes, H and W are prediction dimensions
num_classes = 4
pred_height = 16
pred_width = 20
result = torch.randn(num_classes, pred_height, pred_width)

# Image size that the segmentor took as input (height, width)
# This should be smaller than or equal to pred_height and pred_width
img_size = (12, 15)

# Desired output resolution
output_height = 24
output_width = 32

ARGS = {
    "result": result,
    "img_size": img_size,
    "output_height": output_height,
    "output_width": output_width,
}
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 (4, 24, 32) — shape (4, 24, 32) 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
print_scene_recur Not yet run
unverified (harvested, no recorded run)
jinlinyi/SparsePlanes/data_preparation/utils.py
code served (permissive licence) · get_code("8117d4e34b80b3ab")
How it ran

no invocation recorded.

Also present in other papers' code

Of 18 samples lifted from this paper's linked repositories, 18 carry a function identity and 0 do not. Of the 18 distinct functions with an identity, 5 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 2 the same month, 5 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.

FunctionAlso present inEarliest-datedOthers named
pad_masks 22 other papers
8 dated earlier, 1 the same month, 13 later
1703.06870 (2017-03)
Mask R-CNN
1804.08864 (2018-04)
Learning to See the Invisible: End-to-End Trainable Amodal Instance Segmentation
1908.07836 (2019-08)
PubLayNet: largest dataset ever for document layout analysis
and 19 more not named here
paste_mask_in_image_old 22 other papers
8 dated earlier, 1 the same month, 13 later
1703.06870 (2017-03)
Mask R-CNN
1804.08864 (2018-04)
Learning to See the Invisible: End-to-End Trainable Amodal Instance Segmentation
1908.07836 (2019-08)
PubLayNet: largest dataset ever for document layout analysis
and 19 more not named here
conv2d 2 other papers
2 later
2111.05846 (2021-11)
Structure from Silence: Learning Scene Structure from Ambient Sound
2303.11329 (2023-03)
Sound Localization from Motion: Jointly Learning Sound Direction and Camera Rotation
get_depth_err 1 other paper
1 later
2211.16799 (2022-11)
NOPE-SAC: Neural One-Plane RANSAC for Sparse-View Planar 3D Reconstruction
l1LossMask 1 other paper
1 later
2211.16799 (2022-11)
NOPE-SAC: Neural One-Plane RANSAC for Sparse-View Planar 3D Reconstruction

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

Figure 1: Given two RGB images with an unknown relationship, our system produces a single, coherent planar surface reconstruction of the scene in terms of 3D planes and relative camera poses. Our method succeeds despite the wide baseline and the relatively small amount of overlap in structure (here: 64 • rotation and 1.7m translation and 27% overlap). We show this reconstruction with the inferred left and right cameras in Blue and Red.

For agents

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

get_harvested_code_for_paper("2103.14644")
get_code_for_paper("2103.14644")
have("2103.14644")

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

curl https://syntology.ai/api/ran/2103.14644.json

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

[![Syntology run record](https://syntology.ai/api/ran/2103.14644.svg)](https://syntology.ai/paper/2103.14644)

Connect an agent — have() is free.