SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2003.10983 · 2020

Deep Local Shapes: Learning Local SDF Priors for Detailed 3D Reconstruction

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; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 4 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/2003.10983.json · badge: /api/ran/2003.10983.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
Kamysek/DeepLocalShapes reimplementation (classifier label)
classifier basis: classifier_2025
4 ran · 0 no recorded run
FunctionStatusWhere it lives
filter_classes Ran
ran; our contract draft was wrong, not the code
Kamysek/DeepLocalShapes/preprocess_data.py
code served (permissive licence) · get_code("60567a9af8603a78")
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 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 torch
import fnmatch
import re

# Analyze the function:
# - filter_classes takes two parameters: patterns and classes
# - patterns: a list where first element can be 'glob' or 'regex', followed by pattern strings
# - classes: a list of class name strings (or if patterns[0] is 'glob'/'regex', classes[1:] is used)
# - The function filters class names based on glob or regex patterns

# Test case 1: glob pattern matching
patterns_glob = ['glob', '*.txt', 'test_*']
classes_glob = ['file.txt', 'test_module', 'data.csv', 'test_utils', 'readme.txt']

# Test case 2: regex pattern matching
patterns_regex = ['regex', '^test_.*', '.*\\.py$']
classes_regex = ['test_module', 'test_utils', 'script.py', 'data.csv', 'helper.py']

# Test case 3: default glob (no 'glob' or 'regex' prefix)
patterns_default = ['*.py', 'test_*']
classes_default = ['main.py', 'test_file', 'utils.py', 'readme.txt']

ARGS = {
    'patterns': patterns_glob,
    'classes': classes_glob
}
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 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
filter_classes_glob Ran
ran; our contract draft was wrong, not the code
Kamysek/DeepLocalShapes/preprocess_data.py
code served (permissive licence) · get_code("84cab349d58d0f04")
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 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 fnmatch

ARGS = {
    "patterns": ["test_*", "class_?", "sample*"],
    "classes": ["test_a", "test_b", "class_1", "class_2", "sample_x", "other_class", "test_c"]
}
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 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
filter_classes_regex Ran
ran; our contract draft was wrong, not the code
Kamysek/DeepLocalShapes/preprocess_data.py
code served (permissive licence) · get_code("3a7f1a26d6421562")
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 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 re

# Analyze the function:
# - patterns: iterable of regex pattern strings
# - classes: iterable of class names (strings) to filter
# The function compiles each pattern and filters classes using regex.match

ARGS = {
    "patterns": [r"^Conv", r"Linear.*"],
    "classes": ["Conv2d", "ConvTranspose2d", "Linear", "LinearLayer", "BatchNorm2d", "ReLU"]
}
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 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_learning_rate_schedules Ran
ran; our contract draft was wrong, not the code
Kamysek/DeepLocalShapes/train_deep_ls.py
code served (permissive licence) · get_code("52a0dcc56e4d0c4f")
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 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.

# Define the learning rate schedule classes
class LearningRateSchedule:
    def get_learning_rate(self, epoch):
        pass

class ConstantLearningRateSchedule(LearningRateSchedule):
    def __init__(self, value):
        self.value = value

    def get_learning_rate(self, epoch):
        return self.value

class StepLearningRateSchedule(LearningRateSchedule):
    def __init__(self, initial, interval, factor):
        self.initial = initial
        self.interval = interval
        self.factor = factor

    def get_learning_rate(self, epoch):
        return self.initial * self.factor ** (epoch // self.interval)

class WarmupLearningRateSchedule(LearningRateSchedule):
    def __init__(self, initial, warmed_up, length):
        self.initial = initial
        self.warmed_up = warmed_up
        self.length = length

    def get_learning_rate(self, epoch):
        if epoch > self.length:
            return self.warmed_up
        return self.initial + (self.warmed_up - self.initial) * epoch / self.length

def get_learning_rate_schedules(specs):
    schedule_specs = specs['LearningRateSchedule']
    schedules = []
    for schedule_specs in schedule_specs:
        if schedule_specs['Type'] == 'Step':
            schedules.append(StepLearningRateSchedule(schedule_specs['Initial'], schedule_specs['Interval'], schedule_specs['Factor']))
        elif schedule_specs['Type'] == 'Warmup':
            schedules.append(WarmupLearningRateSchedule(schedule_specs['Initial'], schedule_specs['Final'], schedule_specs['Length']))
        elif schedule_specs['Type'] == 'Constant':
            schedules.append(ConstantLearningRateSchedule(schedule_specs['Value']))
        else:
            raise Exception('no known learning rate schedule of type "{}"'.format(schedule_specs['Type']))
    return schedules

# Construct valid arguments for get_learning_rate_schedules
ARGS = {
    'specs': {
        'LearningRateSchedule': [
            {
                'Type': 'Constant',
                'Value': 0.001
            },
            {
                'Type': 'Step',
                'Initial': 0.01,
                'Interval': 10,
                'Factor': 0.5
            },
            {
                'Type': 'Warmup',
                'Initial': 0.0001,
                'Final': 0.001,
                'Length': 5
            }
        ]
    }
}
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 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

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, 1 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 1 later) and 3 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
get_learning_rate_schedules 2 other papers
2 later
2207.06283 (2022-07)
Implicit Neural Representations for Generative Modeling of Living Cell Shapes
2410.19680 (2024-10)
Inferring Neural Signed Distance Functions by Overfitting on Single Noisy Point Clouds through Finetuning Data-Driven based Priors

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

Efficiently reconstructing complex and intricate surfaces at scale is a long-standing goal in machine perception. To address this problem we introduce Deep Local Shapes (DeepLS), a deep shape representation that enables encoding and reconstruction of high-quality 3D shapes without prohibitive memory requirements. DeepLS replaces the dense volumetric signed distance function (SDF) representation used in traditional surface reconstruction systems with a set of locally learned continuous SDFs defined by a neural network, inspired by recent work such as DeepSDF. Unlike DeepSDF, which represents an object-level SDF with a neural network and a single latent code, we store a grid of independent latent codes, each responsible for storing information about surfaces in a small local neighborhood. This decomposition of scenes into local shapes simplifies the prior distribution that the network must learn, and also enables efficient inference. We demonstrate the effectiveness and generalization power of DeepLS by showing object shape encoding and reconstructions of full scenes, where DeepLS delivers high compression, accuracy, and local shape completion.

For agents

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

get_harvested_code_for_paper("2003.10983")
get_code_for_paper("2003.10983")
have("2003.10983")

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

curl https://syntology.ai/api/ran/2003.10983.json

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

[![Syntology run record](https://syntology.ai/api/ran/2003.10983.svg)](https://syntology.ai/paper/2003.10983)

Connect an agent — have() is free.