SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2007.10453 · 2020

Points2Surf: Learning Implicit Surfaces from Point Cloud Patches

arXiv · PDF · Open in the Atlas

Code that ran

9 of 11 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 6 have no contract verdict; 3 failures of our instrument; 2 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/2007.10453.json · badge: /api/ran/2007.10453.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
ErlerPhilipp/points2surf canonical (classifier label)
classifier basis: classifier_2025
3 ran · 1 no recorded run
MChaus/NeoRender_test_task community (classifier label)
classifier basis not recorded on the link
6 ran · 1 no recorded run
FunctionStatusWhere it lives
decode_sdf Ran
ran on a synthesized input
MChaus/NeoRender_test_task/source/utils.py
code served (permissive licence) · get_code("915eaec694cc6e1c")
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 float, shape (2, 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
import torch.nn as nn

torch.manual_seed(0)

# Create a simple decoder that takes concatenated input and outputs SDF values
class SDFDecoder(nn.Module):
    def __init__(self, input_dim, output_dim=1):
        super().__init__()
        self.linear = nn.Linear(input_dim, output_dim)
    
    def forward(self, x):
        return self.linear(x)

# Set up dimensions with distinct sizes for clarity
latent_dim = 4
query_dim = 3
num_queries = 2

# Create decoder
decoder = SDFDecoder(input_dim=latent_dim + query_dim, output_dim=1)

# Create latent vector (batch_size=1, latent_dim=4)
latent_vector = torch.randn(1, latent_dim)

# Create query points (num_samples=2, query_dim=3)
queries = torch.randn(num_queries, query_dim)

ARGS = {
    "decoder": decoder,
    "latent_vector": latent_vector,
    "queries": queries
}
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 2, dtype family float, shape (2, 1) — shape (2, 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
get_models_path Ran
ran on a synthesized input
MChaus/NeoRender_test_task/sample_points.py
code served (permissive licence) · get_code("e04b438843919fb2")
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 list, ndim 1, 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 glob
import tempfile
import os

# Create a temporary directory with some mock model files
temp_dir = tempfile.mkdtemp()

# Create some dummy model files matching a pattern
os.makedirs(os.path.join(temp_dir, "subdir"), exist_ok=True)
open(os.path.join(temp_dir, "model_v1.pt"), "w").close()
open(os.path.join(temp_dir, "model_v2.pt"), "w").close()
open(os.path.join(temp_dir, "subdir", "model_v3.pt"), "w").close()

# Define pattern to match model files recursively
pattern = os.path.join(temp_dir, "**", "model_*.pt")

ARGS = {
    "pattern": pattern
}
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 list, ndim 1, dtype family float — a list 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_output_dimensions Ran
ran; our contract draft was wrong, not the code
ErlerPhilipp/points2surf/source/points_to_surf_eval.py
code served (permissive licence) · get_code("c2dc2d8c43635f44")
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 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

class TrainOpt:
    def __init__(self, outputs):
        self.outputs = outputs

# Test case 1: all output types
ARGS = {
    'train_opt': TrainOpt(outputs=[
        'imp_surf',
        'imp_surf_magnitude',
        'imp_surf_sign',
        'p_index',
        'patch_pts_ids'
    ])
}
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 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
get_output_ids Ran
ran; our contract draft was wrong, not the code
ErlerPhilipp/points2surf/source/points_to_surf_eval.py
code served (permissive licence) · get_code("5766c560030f7164")
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 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
import numpy as np

class TrainOpt:
    def __init__(self, outputs):
        self.outputs = outputs

# Create a train_opt object with various output types
train_opt = TrainOpt(outputs=[
    'imp_surf',
    'imp_surf_magnitude',
    'imp_surf_sign',
    'p_index',
    'patch_pts_ids',
    'imp_surf',
    'other_output'
])

ARGS = {
    'train_opt': train_opt
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy, 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
get_results_dir Ran
ran on a synthesized input
MChaus/NeoRender_test_task/continue_train.py
code served (permissive licence) · get_code("6f25e98522278446")
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 os
import tempfile
import shutil

# Create a temporary directory for testing
temp_dir = tempfile.mkdtemp()

# Clean up function to remove test directories
def cleanup():
    if os.path.exists(temp_dir):
        shutil.rmtree(temp_dir)

# Change to temp directory so we don't pollute the actual filesystem
original_cwd = os.getcwd()
os.chdir(temp_dir)

# Create the training/DeepSDF base directories
os.makedirs('training/DeepSDF', exist_ok=True)

ARGS = {
    'results_dir': 'test_results'
}
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
laod_train_paths Ran
ran on a synthesized input
MChaus/NeoRender_test_task/continue_train.py
code served (permissive licence) · get_code("70c418780017e23a")
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 list, ndim 1, 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 json
import tempfile
import os

# Create a temporary JSON file with the expected structure
temp_dir = tempfile.mkdtemp()
split_file = os.path.join(temp_dir, "split.json")

data_dict = {
    "train": ["path/to/train1.txt", "path/to/train2.txt", "path/to/train3.txt"],
    "val": ["path/to/val1.txt"],
    "test": ["path/to/test1.txt"]
}

with open(split_file, 'w') as f:
    json.dump(data_dict, f)

ARGS = {
    "split_path": split_file
}
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 list, ndim 1, dtype family float — a list 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
load_classes Ran
ran on a synthesized input
MChaus/NeoRender_test_task/load_data_sets.py
code served (permissive licence) · get_code("4339938210066afc")
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 list, ndim 1, 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 json
import tempfile
import os

# Create a temporary JSON file with class names
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, 'classes.json')

classes_data = {
    'classes': ['cat', 'dog', 'bird', 'fish']
}

with open(temp_file, 'w') as f:
    json.dump(classes_data, f)

ARGS = {
    'filename': temp_file
}
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 list, ndim 1, dtype family float — a list 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
marching_cubes Ran
ran on a synthesized input
MChaus/NeoRender_test_task/source/utils.py
code served (permissive licence) · get_code("d0b9acfcba16fcca")
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 a minimal decoder that takes concatenated latent + query vectors
# and outputs SDF values
class SDFDecoder(nn.Module):
    def __init__(self, latent_dim=8, query_dim=3):
        super().__init__()
        self.linear1 = nn.Linear(latent_dim + query_dim, 16)
        self.linear2 = nn.Linear(16, 1)
    
    def forward(self, x):
        # x is [batch_size, latent_dim + query_dim]
        x = torch.relu(self.linear1(x))
        x = self.linear2(x)
        return x

# Create decoder instance
decoder = SDFDecoder(latent_dim=8, query_dim=3)

# Create latent vector
latent_vec = torch.randn(1, 8)

# Small N value for testing (2 instead of 256 to keep it tiny)
N = 2

ARGS = {
    "decoder": decoder,
    "latent_vec": latent_vec,
    "level": 0.0,
    "N": N,
    "max_batch": 8
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
skimage, 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
parse_arguments Ran
ran; our contract draft was wrong, not the code
ErlerPhilipp/points2surf/source/points_to_surf_eval.py
code served (permissive licence) · get_code("f51709e1f7c95b06")
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 Namespace. 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 argparse

ARGS = {
    'args': None,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 Namespace — a Namespace output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Namespace 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
parse_arguments Not yet run
unverified (harvested, no recorded run)
ErlerPhilipp/points2surf/source/points_to_surf_train.py
code served (permissive licence) · get_code("1cd5046a0ace2eb4")
How it ran

no invocation recorded.

prepare_dir Not yet run
unverified (harvested, no recorded run)
MChaus/NeoRender_test_task/sample_points.py
code served (permissive licence) · get_code("7465b4628f62ca9a")
How it ran

no invocation recorded.

Also present in other papers' code

Of 11 samples lifted from this paper's linked repositories, 11 carry a function identity and 0 do not. Of the 11 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 11 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 2026-11-22.

Repositories linked to this paper

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

Abstract

A key step in any scanning-based asset creation workflow is to convert unordered point clouds to a surface. Classical methods (e.g., Poisson reconstruction) start to degrade in the presence of noisy and partial scans. Hence, deep learning based methods have recently been proposed to produce complete surfaces, even from partial scans. However, such data-driven methods struggle to generalize to new shapes with large geometric and topological variations. We present Points2Surf, a novel patch-based learning framework that produces accurate surfaces directly from raw scans without normals. Learning a prior over a combination of detailed local patches and coarse global information improves generalization performance and reconstruction accuracy. Our extensive comparison on both synthetic and real data demonstrates a clear advantage of our method over state-of-the-art alternatives on previously unseen classes (on average, Points2Surf brings down reconstruction error by 30% over SPR and by 270%+ over deep learning based SotA methods) at the cost of longer computation times and a slight increase in small-scale topological noise in some cases. Our source code, pre-trained model, and dataset are available on: https://github.com/ErlerPhilipp/points2surf

For agents

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

get_harvested_code_for_paper("2007.10453")
get_code_for_paper("2007.10453")
have("2007.10453")

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

curl https://syntology.ai/api/ran/2007.10453.json

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

[![Syntology run record](https://syntology.ai/api/ran/2007.10453.svg)](https://syntology.ai/paper/2007.10453)

Connect an agent — have() is free.