SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2003.03134 · 2020

Bundle Adjustment on a Graph Processor

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; 0 honoured the contract we drafted, 0 violated it and 8 have no contract verdict; 0 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/2003.03134.json · badge: /api/ran/2003.03134.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
joeaortiz/gbp canonical (classifier label)
classifier basis: census
8 ran · 2 no recorded run
FunctionStatusWhere it lives
S03_hat_operator Ran
ran on a synthesized input
joeaortiz/gbp/utils/lie_algebra.py
code served (permissive licence) · get_code("1caf4e2b5ec0da6d")
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 (3, 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

def S03_hat_operator(x):
    """
        Hat operator for SO(3) Lie Group
    """
    return np.array([[0.0, -x[2], x[1]], [x[2], 0.0, -x[0]], [-x[1], x[0], 0.0]])

# The function takes a single parameter x which should be array-like with 3 elements
# It accesses x[0], x[1], x[2] to construct a 3x3 skew-symmetric matrix
x_input = np.array([1.5, 2.3, 0.7])

ARGS = {
    'x': x_input
}
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 (3, 3) — shape (3, 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
SE3_hat_operator Ran
ran on a synthesized input
joeaortiz/gbp/utils/lie_algebra.py
code served (permissive licence) · get_code("0913e8f65172eb6c")
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 (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

# The function SE3_hat_operator takes a minimal representation vector x
# with 6 elements: first 3 for translation, last 3 for rotation
# It constructs a 4x4 matrix for the SE(3) Lie group hat operator

x = np.array([1.0, 2.0, 3.0, 0.5, 0.6, 0.7])

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 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
jac_fn Ran
ran on a synthesized input
joeaortiz/gbp/gbp/factors/linear_displacement.py
code served (permissive licence) · get_code("3f2c29ab00178d2d")
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, 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 jac_fn(x):
    return np.hstack((-np.eye(int(len(x) / 2)), np.eye(int(len(x) / 2))))

# Create a valid input for jac_fn
# The function expects x to be an array-like object with even length
# It creates an identity matrix of size len(x)/2 and horizontally stacks
# the negative and positive versions
x = np.array([1.0, 2.0, 3.0, 4.0])

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 2, dtype family float, shape (2, 4) — shape (2, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
meas_fn Ran
ran on a synthesized input
joeaortiz/gbp/gbp/factors/linear_displacement.py
code served (permissive licence) · get_code("c5aa24afb4dd4258")
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

def meas_fn(x):
    J = np.hstack((-np.eye(int(len(x) / 2)), np.eye(int(len(x) / 2))))
    return J @ x

# Create a small test vector of length 4 (even number required by the function)
x = np.array([1.0, 2.0, 3.0, 4.0])

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 (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
proj Ran
ran on a synthesized input
joeaortiz/gbp/utils/transformations.py
code served (permissive licence) · get_code("b8744491a5e984d3")
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

# Test case 1: 1D array
x_1d = np.array([1.0, 2.0, 3.0, 4.0])

# Test case 2: 2D array
x_2d = np.array([
    [1.0, 2.0, 3.0, 4.0],
    [5.0, 6.0, 7.0, 8.0]
])

ARGS = {
    "x": x_1d
}
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
proj_derivative Ran
ran on a synthesized input
joeaortiz/gbp/utils/derivatives.py
code served (permissive licence) · get_code("5fa7ef9ce95273fd")
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 (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 numpy as np

# Create a 1D numpy array to pass to proj_derivative
# The function checks if x.ndim == 1 and uses x[-1] as a divisor
# and x[:-1] for slicing, so we need at least 2 elements
x = np.array([1.0, 2.0, 3.0, 4.0])

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 2, dtype family float, shape (3, 4) — shape (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
read_balfile Ran
ran on a synthesized input
joeaortiz/gbp/utils/read_balfile.py
code served (permissive licence) · get_code("1150d38034b9542c")
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 numpy as np
import tempfile
import os

# Create a temporary .bal file with valid content
temp_dir = tempfile.mkdtemp()
balfile_path = os.path.join(temp_dir, "test.bal")

# Write a valid .bal file
with open(balfile_path, 'w') as f:
    # Header comment
    f.write("# This is a comment\n")
    
    # n_keyframes, n_points, n_edges
    n_keyframes = 2
    n_points = 3
    n_edges = 4
    f.write(f"{n_keyframes} {n_points} {n_edges}\n")
    
    # Camera intrinsic matrix K: fx fy cx cy
    f.write("500.0 500.0 320.0 240.0\n")
    
    # Measurements (n_edges lines): cam_id point_id x y
    for i in range(n_edges):
        f.write(f"{i % n_keyframes} {i % n_points} {100.0 + i} {200.0 + i}\n")
    
    # Camera means (n_keyframes * 6 lines): one value per line
    for i in range(n_keyframes * 6):
        f.write(f"{0.1 * i}\n")
    
    # Landmark means (n_points * 3 lines): one value per line
    for i in range(n_points * 3):
        f.write(f"{0.2 * i}\n")

ARGS = {
    "balfile": balfile_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
so3exp Ran
ran on a synthesized input
joeaortiz/gbp/utils/lie_algebra.py
code served (permissive licence) · get_code("2029a9e8c0de57f6")
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 (3, 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
import torch

torch.manual_seed(0)

# The function so3exp takes a single parameter 'w' which should be a vector
# representing an element of so(3) (the Lie algebra of SO(3)).
# Looking at the function body:
# - w is used with np.linalg.norm(w) to compute theta
# - w is passed to S03_hat_operator(w) which indexes it as x[0], x[1], x[2]
# - So w should be a 1D array-like object with at least 3 elements

# Create a small valid input for so(3) - a 3-element vector
w = np.array([0.1, 0.2, 0.3])

ARGS = {
    "w": w
}
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 ndarray, ndim 2, dtype family float, shape (3, 3) — shape (3, 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
getT_qt Not yet run
unverified (harvested, no recorded run)
joeaortiz/gbp/utils/transformations.py
code served (permissive licence) · get_code("842b6718320b3d40")
How it ran

no invocation recorded.

jac_fd Not yet run
unverified (harvested, no recorded run)
joeaortiz/gbp/utils/derivatives.py
code served (permissive licence) · get_code("0cd1741da6f7b3c9")
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, 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 10 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 2026-11-21.

Repositories linked to this paper

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

Abstract

Graph processors such as Graphcore's Intelligence Processing Unit (IPU) are part of the major new wave of novel computer architecture for AI, and have a general design with massively parallel computation, distributed on-chip memory and very high inter-core communication bandwidth which allows breakthrough performance for message passing algorithms on arbitrary graphs. We show for the first time that the classical computer vision problem of bundle adjustment (BA) can be solved extremely fast on a graph processor using Gaussian Belief Propagation. Our simple but fully parallel implementation uses the 1216 cores on a single IPU chip to, for instance, solve a real BA problem with 125 keyframes and 1919 points in under 40ms, compared to 1450ms for the Ceres CPU library. Further code optimisation will surely increase this difference on static problems, but we argue that the real promise of graph processing is for flexible in-place optimisation of general, dynamically changing factor graphs representing Spatial AI problems. We give indications of this with experiments showing the ability of GBP to efficiently solve incremental SLAM problems, and deal with robust cost functions and different types of factors.

For agents

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

get_harvested_code_for_paper("2003.03134")
get_code_for_paper("2003.03134")
have("2003.03134")

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

curl https://syntology.ai/api/ran/2003.03134.json

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

[![Syntology run record](https://syntology.ai/api/ran/2003.03134.svg)](https://syntology.ai/paper/2003.03134)

Connect an agent — have() is free.