SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 136990210 · ECCV · 2022

FH-Net: A Fast Hierarchical Network for Scene Flow Estimation on Real-world Point Clouds

Jie Wang, Tingfa Xu, Lihe Ding, Jianan Li, Xinli Xu, Shaocong Dong

Open in the Atlas

Code that ran

10 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; 5 honoured the contract we drafted, 0 violated it and 5 have no contract verdict; 0 failures of our instrument; 0 no recorded run), measured 2026-11-22; 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/136990210.json · badge: /api/ran/136990210.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
pigtigger/FH-Net canonical (classifier label)
classifier basis: census
10 ran · 0 no recorded run
FunctionStatusWhere it lives
box_camera_to_lidar Ran
ran, honoured the contract we drafted
pigtigger/FH-Net/utils/data_utils/box_np_ops.py
code served (permissive licence) · get_code("96c8084488bd023b")
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, 7) 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

np.random.seed(0)

# Create small sample data for box_camera_to_lidar function
# data shape should be [N, 7] where N is number of boxes
# Each box has: [x, y, z, l, h, w, r] (3 position + 3 dimensions + 1 rotation)
data = np.array([
    [5.0, 10.0, 2.0, 3.5, 1.6, 1.8, 0.5],
    [8.0, 12.0, 1.5, 3.2, 1.5, 1.7, 1.2]
], dtype=np.float32)

# r_rect: [4, 4] matrix to project from CAM2 to CAM0
# This is typically the camera rectification matrix
r_rect = np.array([
    [0.9999, -0.0050, 0.0095, 0.0],
    [0.0050, 0.9999, -0.0010, 0.0],
    [-0.0095, 0.0010, 0.9999, 0.0],
    [0.0, 0.0, 0.0, 1.0]
], dtype=np.float32)

# velo2cam: [4, 4] matrix to project from lidar/velodyne to camera coordinate
# This is the extrinsic calibration matrix
velo2cam = np.array([
    [0.0125, -0.9992, 0.0359, -0.0430],
    [-0.2357, -0.0381, -0.9714, -0.2330],
    [0.9719, 0.0106, -0.2360, -0.0144],
    [0.0, 0.0, 0.0, 1.0]
], dtype=np.float32)

ARGS = {
    'data': data,
    'r_rect': r_rect,
    'velo2cam': velo2cam
}
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, 7) — shape (2, 7) 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
camera_to_lidar Ran
ran, honoured the contract we drafted
pigtigger/FH-Net/utils/data_utils/box_np_ops.py
code served (permissive licence) · get_code("9e22984820c39868")
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

# Create small test data for camera_to_lidar function
# points: shape [N, 3] - points in camera coordinate
np.random.seed(0)
points = np.random.randn(2, 3).astype(np.float32)

# r_rect: shape [4, 4] - matrix to project points from specific camera coord to CAM0
# Create a valid 4x4 transformation matrix
r_rect = np.eye(4, dtype=np.float32)
r_rect[0, 1] = 0.1
r_rect[1, 2] = 0.05

# velo2cam: shape [4, 4] - matrix to project points from camera coord to lidar coord
# Create a valid 4x4 transformation matrix
velo2cam = np.eye(4, dtype=np.float32)
velo2cam[0, 0] = 0.9
velo2cam[1, 1] = 0.95
velo2cam[2, 2] = 1.0

ARGS = {
    "points": points,
    "r_rect": r_rect,
    "velo2cam": velo2cam,
}
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
compute_gt Ran
ran on a synthesized input
pigtigger/FH-Net/utils/data_utils/kitti_utils.py
code served (permissive licence) · get_code("eab7ec48c1999700")
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
import numpy as np

torch.manual_seed(0)

# Create small point clouds in box coordinates (N x 4 for homogeneous coords)
pc1_box = torch.randn(2, 4, dtype=torch.float32)
pc2_box = torch.randn(2, 4, dtype=torch.float32)

# Create small point clouds in camera coordinates (N x 4)
pc1_cam = torch.randn(2, 4, dtype=torch.float32)
pc2_cam = torch.randn(2, 4, dtype=torch.float32)

# Create transformation matrices (4 x 4 homogeneous transforms)
T1_box2cam = torch.randn(4, 4, dtype=torch.float32)
T2_box2cam = torch.randn(4, 4, dtype=torch.float32)

ARGS = {
    'pc1_cam': pc1_cam,
    'pc2_cam': pc2_cam,
    'pc1_box': pc1_box,
    'pc2_box': pc2_box,
    'T1_box2cam': T1_box2cam,
    'T2_box2cam': T2_box2cam
}
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 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
get_T_box_to_lidar Ran
ran on a synthesized input
pigtigger/FH-Net/utils/data_utils/box_np_ops.py
code served (permissive licence) · get_code("a669c02d62b4891e")
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 (2, 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

# Create sample data: N boxes with 7 values each [x, y, z, width, length, height, rotation_z]
num_boxes = 2
data = np.array([
    [1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 0.5],
    [7.0, 8.0, 9.0, 10.0, 11.0, 12.0, 1.2]
], dtype=np.float32)

ARGS = {
    "data": data
}
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 (2, 4, 4) — shape (2, 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
get_image_path Ran
ran on a synthesized input
pigtigger/FH-Net/utils/data_utils/waymo_utils.py
code served (permissive licence) · get_code("76345343c491493c")
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 str. 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 tempfile
from pathlib import Path

# Create a temporary directory structure that mimics KITTI dataset
temp_dir = tempfile.mkdtemp()
prefix = temp_dir

# Create the necessary directory structure
scene_dir = Path(prefix) / "000"
image_dir = scene_dir / "image_2"
image_dir.mkdir(parents=True, exist_ok=True)

# Create a dummy image file
idx = 5
filename = f"{idx:07d}.png"
dummy_image_path = image_dir / filename
dummy_image_path.touch()

# Define arguments for get_image_path
ARGS = {
    'idx': 5,
    'prefix': prefix,
    'relative_path': True,
    'exist_check': True,
    'info_type': 'image_2'
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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_kitti_info_path Ran
ran on a synthesized input
pigtigger/FH-Net/utils/data_utils/waymo_utils.py
code served (permissive licence) · get_code("9461f54c8e1b822c")
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 str. 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.

from pathlib import Path
import tempfile
import os

# Create a temporary directory structure to simulate KITTI dataset
temp_dir = tempfile.mkdtemp()
scene_dir = Path(temp_dir) / "000"
scene_dir.mkdir(parents=True, exist_ok=True)
info_type_dir = scene_dir / "info_type"
info_type_dir.mkdir(parents=True, exist_ok=True)

# Create a dummy file
dummy_file = info_type_dir / "0000042.txt"
dummy_file.touch()

ARGS = {
    'idx': 42,
    'prefix': temp_dir,
    'info_type': 'info_type',
    'file_tail': '.txt',
    'relative_path': False,
    'exist_check': True,
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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_label_path Ran
ran on a synthesized input
pigtigger/FH-Net/utils/data_utils/waymo_utils.py
code served (permissive licence) · get_code("efd5b6e47d7184be")
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 str. 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.

from pathlib import Path
import tempfile
import os

# Create a temporary directory structure that mimics KITTI dataset layout
temp_dir = tempfile.mkdtemp()
prefix = temp_dir

# Create the necessary directory structure
# For idx=5, scene_path will be '000' (5 // 1000 % 1000)
# For idx=1005, scene_path will be '001' (1005 // 1000 % 1000)
scene_path_000 = Path(prefix) / '000' / 'label_2'
scene_path_000.mkdir(parents=True, exist_ok=True)

# Create a dummy label file for idx=5
dummy_label_file = scene_path_000 / '0000005.txt'
dummy_label_file.write_text('dummy label data')

ARGS = {
    'idx': 5,
    'prefix': prefix,
    'relative_path': False,
    'exist_check': True,
    'info_type': 'label_2'
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
rotx Ran
ran, honoured the contract we drafted
pigtigger/FH-Net/utils/data_utils/kitti_oxts.py
code served (permissive licence) · get_code("7461efe8005a3417")
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 rotx(t):
    """Rotation about the x-axis."""
    c = np.cos(t)
    s = np.sin(t)
    return np.array([[1, 0, 0], [0, c, -s], [0, s, c]])

# The function takes a single parameter t (rotation angle in radians)
# and returns a 3x3 rotation matrix about the x-axis.
# t should be a scalar (float or numpy scalar).

ARGS = {
    "t": 0.5  # A small rotation angle in radians
}
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
roty Ran
ran, honoured the contract we drafted
pigtigger/FH-Net/utils/data_utils/kitti_oxts.py
pointer only (licence: MIT); the stored copy get_code serves, from tingxueronghua/dpke, is pointer only (licence: NONE) · get_code("e1f78f900f99690f")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from tingxueronghua/dpke, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

rotz Ran
ran, honoured the contract we drafted
pigtigger/FH-Net/utils/data_utils/kitti_oxts.py
code served (permissive licence) · get_code("090a666d56e7bca7")
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 rotz(t):
    """Rotation about the z-axis."""
    c = np.cos(t)
    s = np.sin(t)
    return np.array([[c, -s, 0], [s, c, 0], [0, 0, 1]])

# The function takes a single parameter t (angle in radians)
# and returns a 3x3 rotation matrix about the z-axis
# t should be a scalar or array-like numeric value

t_value = np.pi / 4  # 45 degrees as a simple test value

ARGS = {
    "t": t_value
}
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

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, 5 are also present in other papers whose code we hold (5 with at least one paper dated earlier, 0 the same month, 4 later) and 5 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
roty 13 other papers
6 dated earlier, 3 later, 4 same year, order not determinable
1612.07695 (2016-12)
MultiNet: Real-time Joint Semantic Reasoning for Autonomous Driving
1811.07782 (2018-11)
Modeling Local Geometric Structure of 3D Point Clouds using Geo-CNN
1912.12147 (2019-12)
Cooperative Perception for 3D Object Detection in Driving Scenarios using Infrastructure Sensors
and 10 more not named here
rotx 12 other papers
6 dated earlier, 2 later, 4 same year, order not determinable
1612.07695 (2016-12)
MultiNet: Real-time Joint Semantic Reasoning for Autonomous Driving
1811.07782 (2018-11)
Modeling Local Geometric Structure of 3D Point Clouds using Geo-CNN
1912.12147 (2019-12)
Cooperative Perception for 3D Object Detection in Driving Scenarios using Infrastructure Sensors
and 9 more not named here
box_camera_to_lidar 9 other papers
1 dated earlier, 5 later, 3 same year, order not determinable
2104.10956 (2021-04)
FCOS3D: Fully Convolutional One-Stage Monocular 3D Object Detection
2203.10642 (2022-03)
FUTR3D: A Unified Sensor Fusion Framework for 3D Detection
2205.13542 (2022-05)
BEVFusion: Multi-Task Multi-Sensor Fusion with Unified Bird's-Eye View Representation
and 6 more not named here
camera_to_lidar 9 other papers
1 dated earlier, 5 later, 3 same year, order not determinable
2104.10956 (2021-04)
FCOS3D: Fully Convolutional One-Stage Monocular 3D Object Detection
2203.10642 (2022-03)
FUTR3D: A Unified Sensor Fusion Framework for 3D Detection
2205.13542 (2022-05)
BEVFusion: Multi-Task Multi-Sensor Fusion with Unified Bird's-Eye View Representation
and 6 more not named here
rotz 8 other papers
6 dated earlier, 2 same year, order not determinable
1612.07695 (2016-12)
MultiNet: Real-time Joint Semantic Reasoning for Autonomous Driving
1811.07782 (2018-11)
Modeling Local Geometric Structure of 3D Point Clouds using Geo-CNN
1912.12147 (2019-12)
Cooperative Perception for 3D Object Detection in Driving Scenarios using Infrastructure Sensors
and 5 more not named here

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

Estimating scene flow from real-world point clouds is a fundamental task for practical 3D vision. Previous methods often rely on deep models to first extract expensive per-point features at full resolution, and then get the flow either from complex matching mechanism or feature decoding, suffering high computational cost and latency. In this work, we propose a fast hierarchical network, FH-Net, which directly gets the key points flow through a lightweight Trans-flow layer utilizing the reliable local geometry prior, and optionally back-propagates the computed sparse flows through an inverse Trans-up layer to obtain hierarchical flows at different resolutions. To focus more on challenging dynamic objects, we also provide a new copy-and-paste data augmentation technique based on dynamic object pairs generation. Moreover, to alleviate the chronic shortage of real-world training data, we establish two new large-scale datasets to this field by collecting lidar-scanned point clouds from public autonomous driving datasets and annotating the collected data through novel pseudo-labeling. Extensive experiments on both public and proposed datasets show that our method outperforms prior state-of-the-arts while running at least 7× faster at 113 FPS. Code and data are released at https://github.com/pigtigger/FH-Net.

For agents

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

get_harvested_code_for_paper("136990210")
get_code_for_paper("136990210")
have("136990210")

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

curl https://syntology.ai/api/ran/136990210.json

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

[![Syntology run record](https://syntology.ai/api/ran/136990210.svg)](https://syntology.ai/paper/136990210)

Connect an agent — have() is free.