SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2403.04739 · ECCV · 2024

I Can't Believe It's Not Scene Flow!

Deva Ramanan, Neehar Peri, James Hays, Kyle Vedder, Ishan Khatri, Stack Av

arXiv · PDF · Open in the Atlas

Code that ran

8 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; 1 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 5 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/2403.04739.json · badge: /api/ran/2403.04739.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
kylevedder/bucketedscenefloweval canonical (classifier label)
classifier basis: census
6 ran · 0 no recorded run
kylevedder/SceneFlowZoo community (classifier label)
classifier basis not recorded on the link
2 ran · 0 no recorded run
FunctionStatusWhere it lives
align_vector_to_another Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
kylevedder/bucketedscenefloweval/bucketed_scene_flow_eval/datastructures/line_mesh.py
code served (permissive licence) · get_code("5c5482d289a07f32")
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

# Create two distinct 3D vectors for alignment
a = np.array([0.0, 0.0, 1.0])
b = np.array([1.0, 0.0, 0.0])

ARGS = {
    "a": a,
    "b": b,
}
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
build_metadata Ran
ran on a synthesized input
kylevedder/SceneFlowZoo/av2_scene_flow_competition_submit.py
code served (permissive licence) · get_code("75b0424130303c2c")
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

torch.manual_seed(0)

ARGS = {
    'is_supervised': 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 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
from_fixed_array Ran
ran, honoured the contract we drafted
kylevedder/bucketedscenefloweval/bucketed_scene_flow_eval/datastructures/pointcloud.py
code served (permissive licence) · get_code("8932de372132539f")
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
import torch

torch.manual_seed(0)

# Test case 1: 2D numpy array with some NaN values
array_2d_np = np.array([
    [1.0, 2.0, 3.0],
    [np.nan, 5.0, 6.0],
    [7.0, 8.0, 9.0],
    [np.nan, 11.0, 12.0]
])

# Test case 2: 1D numpy array with some NaN values
array_1d_np = np.array([1.0, np.nan, 3.0, 4.0, np.nan])

# Test case 3: 2D torch tensor with some NaN values
array_2d_torch = torch.tensor([
    [1.0, 2.0, 3.0],
    [float('nan'), 5.0, 6.0],
    [7.0, 8.0, 9.0]
])

# Test case 4: 1D torch tensor with some NaN values
array_1d_torch = torch.tensor([1.0, float('nan'), 3.0, 4.0])

ARGS = {
    "array": array_2d_np
}
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 (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
load_json Ran
ran on a synthesized input
kylevedder/SceneFlowZoo/data_prep_scripts/flatten_non_causal.py
code served (permissive licence) · get_code("a9e13d8e6c527e20")
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
from pathlib import Path
import tempfile
import os

# Create a temporary JSON file with valid content
temp_dir = tempfile.mkdtemp()
temp_file_path = Path(temp_dir) / "test.json"

test_data = {
    "key1": 42,
    "key2": 100,
    "key3": 7
}

with temp_file_path.open('w') as f:
    json.dump(test_data, f)

ARGS = {
    "file_path": temp_file_path
}
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
make_image_pixel_coordinate_grid Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
kylevedder/bucketedscenefloweval/bucketed_scene_flow_eval/datastructures/pointcloud.py
code served (permissive licence) · get_code("6fa9a0952102a0c3")
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 (12, 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

# The function takes image_shape as a parameter
# image_shape must be a 2-tuple of integers representing (height, width)
# Let's create a small image shape for testing

image_shape = (4, 3)  # height=4, width=3

ARGS = {
    'image_shape': image_shape
}
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 (12, 2) — shape (12, 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
normalized Ran
ran; our contract draft was wrong, not the code
kylevedder/bucketedscenefloweval/bucketed_scene_flow_eval/datastructures/line_mesh.py
code served (permissive licence) · get_code("a9bf78961f9edbd5")
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

# Create a small numpy array of points to normalize
# Using shape (4, 3) - 4 points in 3D space
a = np.array([
    [1.0, 0.0, 0.0],
    [3.0, 4.0, 0.0],
    [0.0, 0.0, 0.0],
    [1.0, 1.0, 1.0]
], dtype=np.float32)

ARGS = {
    "a": a,
    "axis": -1,
    "order": 2
}
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
ray_triangle_intersect Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
kylevedder/bucketedscenefloweval/flow_lab/flow_lab.py
code served (permissive licence) · get_code("76dac6858417dbfb")
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

# Ray-triangle intersection test
# The function expects:
# - ray_origin: 3D point (numpy array)
# - ray_direction: 3D direction vector (numpy array)
# - v0, v1, v2: triangle vertices as 3D points (numpy arrays)

# Create a simple test case with a ray that intersects a triangle
ray_origin = np.array([0.0, 0.0, -5.0])
ray_direction = np.array([0.0, 0.0, 1.0])

# Define a triangle in the XY plane at z=0
v0 = np.array([-1.0, -1.0, 0.0])
v1 = np.array([1.0, -1.0, 0.0])
v2 = np.array([0.0, 1.0, 0.0])

ARGS = {
    'ray_origin': ray_origin,
    'ray_direction': ray_direction,
    'v0': v0,
    '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 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
to_fixed_array Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
kylevedder/bucketedscenefloweval/bucketed_scene_flow_eval/datastructures/pointcloud.py
code served (permissive licence) · get_code("fb9040070a43fabf")
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 (5, 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

# Case 1: Array shorter than max_len - will be padded
array_short = np.array([[1.0, 2.0, 3.0], [4.0, 5.0, 6.0]])
max_len_short = 5
pad_val_short = np.nan

# Case 2: Array longer than max_len - will be sliced
array_long = np.array([[i, i+1, i+2] for i in range(10)])
max_len_long = 3
pad_val_long = np.nan

# Use the short case for the primary ARGS
ARGS = {
    "array": array_short,
    "max_len": max_len_short,
    "pad_val": pad_val_short
}
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 (5, 3) — shape (5, 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

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, 1 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 0 later) and 7 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
normalized 2 other papers
2 dated earlier
1902.04615 (2019-02)
Gauge Equivariant Convolutional Networks and the Icosahedral CNN
2303.01932 (2023-03)
MobileBrick: Building LEGO for 3D Reconstruction on Mobile Devices

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

Georgia Tech (a) Ground Truth (b) FastFlow3D [12] (c) DeFlow [53] (d) NSFP [18] (e) ZeroFlow XL 5x [42] (f ) TrackFlow (Ours) Fig. 1: We visualize an example of two pedestrians (walking from left to right), cherrypicked to have unusually high density lidar returns, making it particularly easy to estimate flow. We expect that state-of-the-art scene flow methods should work well in this case, but find that all prior art fails catastrophically. Notably, TrackFlow is the only method to estimate flow for these pedestrians.

For agents

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

get_harvested_code_for_paper("2403.04739")
get_code_for_paper("2403.04739")
have("2403.04739")

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

curl https://syntology.ai/api/ran/2403.04739.json

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

[![Syntology run record](https://syntology.ai/api/ran/2403.04739.svg)](https://syntology.ai/paper/2403.04739)

Connect an agent — have() is free.