SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2303.17167 · CVPR · 2023

Rethinking the Approximation Error in 3D Surface Fitting for Point Cloud Normal Estimation

Di Xie, Shiliang Pu, Jingjing Wang, Xuejun Yan, Hang Du

arXiv · PDF · Open in the Atlas

Code that ran

4 of 4 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 0 have no contract verdict; 3 failures of our instrument; 0 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2303.17167.json · badge: /api/ran/2303.17167.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
hikvision-research/3dvision canonical (classifier label)
classifier basis: classifier_2025
4 ran · 0 no recorded run
FunctionStatusWhere it lives
fit_Wjet Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
hikvision-research/3dvision/Normal-Estimation/ZTEE/models/AdaFit_ms_HG.py
code served (permissive licence) · get_code("67196e77547ae476")
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
import torch.nn as nn

torch.manual_seed(0)

# Create small tensors for the function arguments
# points: shape (batch_size, D, n_points) where D=3 (x,y,z coordinates)
batch_size = 2
n_points = 25
D = 3

points = torch.randn(batch_size, D, n_points)

# weights: shape (batch_size, n_points) - one weight per point
weights = torch.rand(batch_size, n_points)

# order: polynomial order (1, 2, 3, or 4)
order = 2

# compute_neighbor_normals: boolean flag
compute_neighbor_normals = False

ARGS = {
    "points": points,
    "weights": weights,
    "order": order,
    "compute_neighbor_normals": compute_neighbor_normals,
}
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
knn Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
hikvision-research/3dvision/Normal-Estimation/ZTEE/models/AdaFit_ms_HG.py
code served (permissive licence) · get_code("c779aa8bd2f53710")
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) and returned Tensor, ndim 3, dtype family int, shape (2, 3, 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 torch
import torch.nn as nn

torch.manual_seed(0)

# The function knn expects:
# - x: a tensor that will be transposed (2, 1) and have operations like sum, matmul
# - k: an integer for topk selection

# Looking at the function body:
# - x.transpose(2, 1) suggests x has 3 dimensions
# - torch.sum(x ** 2, dim=1, keepdim=True) operates on dim 1
# - The result is used with topk(k=k, dim=-1)
# This suggests x should be shape (batch, num_points, features)

# Create a small tensor: batch_size=2, num_points=4, features=3
x = torch.randn(2, 4, 3)

# k should be less than the number of points
k = 2

ARGS = {
    "x": x,
    "k": k,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 int, shape (2, 3, 2) — shape (2, 3, 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
l2_norm Ran
ran, honoured the contract we drafted
hikvision-research/3dvision/Normal-Estimation/ZTEE/evaluate.py
code served (permissive licence) · get_code("7f8230b3084eaaab")
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

# The function l2_norm computes the L2 norm of vectors along axis=1
# It expects a 2D array where each row is a vector

# Create a simple 2D numpy array with shape (batch_size, feature_dim)
# Using distinct sizes: 2 samples, 4 features
v = np.array([
    [1.0, 2.0, 3.0, 4.0],
    [5.0, 6.0, 7.0, 8.0]
], dtype=np.float32)

ARGS = {
    "v": v
}
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
solve_linear_system Ran
ran; our contract draft was wrong, not the code
hikvision-research/3dvision/Normal-Estimation/ZTEE/models/AdaFit_ms_HG.py
code served (permissive licence) · get_code("bc24576fff5169fd")
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, 4, 1) 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 valid arguments for solve_linear_system
# The function expects:
# - XtX: a batch of square matrices (shape: [batch_size, n_elements, n_elements])
# - XtY: a batch of vectors (shape: [batch_size, n_elements, 1])
# - sub_batch_size: optional batch size for processing

batch_size = 2
n_elements = 4

# XtX should be positive definite matrices for Cholesky decomposition
# Create symmetric positive definite matrices
XtX = torch.randn(batch_size, n_elements, n_elements)
XtX = torch.matmul(XtX, XtX.transpose(-2, -1)) + torch.eye(n_elements).unsqueeze(0) * 5

# XtY should be a vector of compatible shape
XtY = torch.randn(batch_size, n_elements, 1)

# sub_batch_size should divide batch_size evenly
sub_batch_size = 1

ARGS = {
    'XtX': XtX,
    'XtY': XtY,
    'sub_batch_size': sub_batch_size
}
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 (2, 4, 1) — shape (2, 4, 1) 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 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 distinct functions with an identity, 4 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 1 the same month, 4 later) and 0 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
knn 52 other papers
29 dated earlier, 2 the same month, 21 later · 12 only as an assembly brick
1801.07829 (2018-01)
Dynamic Graph CNN for Learning on Point Clouds
1905.06292 (2019-05)
3D Point Cloud Generative Adversarial Network Based on Tree Structured Graph Convolutions
2005.06734 (2020-05)
Dense-Resolution Network for Point Cloud Classification and Segmentation
and 49 more not named here
fit_Wjet 3 other papers
2 dated earlier, 1 later
2108.05836 (2021-08)
AdaFit: Rethinking Learning-based Normal Estimation on Point Clouds
2211.16869 (2022-11)
NeAF: Learning Neural Angle Fields for Point Normal Estimation
2403.05117 (2024-03)
Arbitrary-Scale Point Cloud Upsampling by Voxel-Based Network with Latent Geometric-Consistent Learning
solve_linear_system 3 other papers
2 dated earlier, 1 later
2108.05836 (2021-08)
AdaFit: Rethinking Learning-based Normal Estimation on Point Clouds
2211.16869 (2022-11)
NeAF: Learning Neural Angle Fields for Point Normal Estimation
2403.05117 (2024-03)
Arbitrary-Scale Point Cloud Upsampling by Voxel-Based Network with Latent Geometric-Consistent Learning
l2_norm 1 other paper
1 later
2403.05117 (2024-03)
Arbitrary-Scale Point Cloud Upsampling by Voxel-Based Network with Latent Geometric-Consistent Learning

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

Most existing approaches for point cloud normal estimation aim to locally fit a geometric surface and calculate the normal from the fitted surface. Recently, learningbased methods have adopted a routine of predicting pointwise weights to solve the weighted least-squares surface fitting problem. Despite achieving remarkable progress, these methods overlook the approximation error of the fitting problem, resulting in a less accurate fitted surface. In this paper, we first carry out in-depth analysis of the approximation error in the surface fitting problem. Then, in order to bridge the gap between estimated and precise surface normals, we present two basic design principles: 1) applies the Z-direction Transform to rotate local patches for a better surface fitting with a lower approximation error; 2) models the error of the normal estimation as a learnable term. We implement these two principles using deep neural networks, and integrate them with the state-of-theart (SOTA) normal estimation methods in a plug-and-play manner. Extensive experiments verify our approaches bring benefits to point cloud normal estimation and push the frontier of state-of-the-art performance on both synthetic and real-world datasets. The code is available at https:// github.com/hikvision-research/3DVision.

For agents

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

get_harvested_code_for_paper("2303.17167")
get_code_for_paper("2303.17167")
have("2303.17167")

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

curl https://syntology.ai/api/ran/2303.17167.json

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

[![Syntology run record](https://syntology.ai/api/ran/2303.17167.svg)](https://syntology.ai/paper/2303.17167)

Connect an agent — have() is free.