SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2006.09043 · 2020

Improved Deep Point Cloud Geometry Compression

arXiv · PDF · Open in the Atlas

Code that ran

15 of 15 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 14 have no contract verdict; 0 failures of our instrument; 0 no recorded run); 1 of them was also run on the host, not in the jail, by the peer-agreement lane that set its level, measured 2026-11-21; 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/2006.09043.json · badge: /api/ran/2006.09043.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
mauriceqch/pcc_geo_cnn_v2 canonical (classifier label)
classifier basis: plan_row
3 ran · 0 no recorded run
pmclSF/DeepCompress community (classifier label)
classifier basis not recorded on the link
12 ran · 0 no recorded run
FunctionStatusWhere it lives
flatten Ran
ran, honoured the contract we drafted
mauriceqch/pcc_geo_cnn_v2/src/ev_experiment.py
code served (permissive licence) · get_code("20922c79a0929bbf")
How it ran

Ran in Syntology's jail on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned list, ndim 1, dtype family int; no output shape is stated. The image it ran in was not recorded. The comparison behind this sample's level is a different run: the peer-agreement lane ran it again on the host, not in the jail (report dated 2026-09-04T22:50:02Z). 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)

# The function flatten takes a list of lists and flattens it into a single list
# It iterates through sublists and extracts items from each

# Create a simple nested list structure
nested_list = [
    [1, 2, 3],
    [4, 5],
    [6, 7, 8, 9]
]

ARGS = {
    "l": nested_list
}
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 list, ndim 1, dtype family int — 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
benchmark_function Ran
ran on a synthesized input
pmclSF/DeepCompress/src/benchmarks.py
code served (permissive licence) · get_code("789e06ac86f87db2")
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 BenchmarkResult. 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 time
from dataclasses import dataclass, field
from typing import Any, Callable, Dict, Optional
import tensorflow as tf

# Define a simple test function to benchmark
def simple_add(a, b):
    """A simple function to benchmark."""
    return a + b

def matrix_multiply(matrix_a, matrix_b):
    """A function that performs matrix multiplication."""
    return tf.matmul(matrix_a, matrix_b)

# Create small test data
test_tensor_1 = tf.constant([[1.0, 2.0], [3.0, 4.0]], dtype=tf.float32)
test_tensor_2 = tf.constant([[5.0, 6.0], [7.0, 8.0]], dtype=tf.float32)

scalar_a = 5
scalar_b = 3

# Define ARGS dictionary with all parameters
ARGS = {
    'func': simple_add,
    'args': (scalar_a, scalar_b),
    'kwargs': {},
    'warmup': 2,
    'iterations': 5,
    'name': 'test_benchmark'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type BenchmarkResult — a BenchmarkResult output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a BenchmarkResult 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
build_points_threshold Ran
ran on a synthesized input
mauriceqch/pcc_geo_cnn_v2/src/model_opt.py
code served (permissive licence) · get_code("9ecb6f04dc82e757")
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. 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 sample data for the function
np.random.seed(0)

# x_hat: array of values that will be compared against thresholds
x_hat = np.array([0.1, 0.5, 0.8, 1.2, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0], dtype='float32')

# thresholds: list of threshold values to test
thresholds = [0.2, 0.7, 1.0, 1.5, 2.5, 3.5, 4.5]

# len_block: reference length for computing len_ratio
len_block = 10

# max_delta: ratio bounds for filtering results
max_delta = 2.0

ARGS = {
    'x_hat': x_hat,
    'thresholds': thresholds,
    'len_block': len_block,
    'max_delta': max_delta
}
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 list — a list 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
compare_implementations Ran
ran on a synthesized input
pmclSF/DeepCompress/src/benchmarks.py
code served (permissive licence) · get_code("752a7e3e54b2c484")
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 time
from dataclasses import dataclass, field
from typing import Any, Callable, Dict, Optional
import tensorflow as tf

# Define simple test functions to benchmark
def add_numbers(a, b):
    """Simple addition function."""
    return a + b

def multiply_numbers(a, b):
    """Simple multiplication function."""
    return a * b

def matrix_operation(matrix):
    """Simple matrix operation."""
    return matrix @ matrix.T

# Set up implementations dict
implementations = {
    "add": add_numbers,
    "multiply": multiply_numbers,
    "matrix_op": matrix_operation,
}

# Set up args and kwargs
args_for_add_multiply = (5, 3)
kwargs_for_add_multiply = {}

# For matrix operation, we need a small matrix
import numpy as np
small_matrix = np.array([[1.0, 2.0, 3.0], [4.0, 5.0, 6.0]], dtype=np.float32)
args_for_matrix = (small_matrix,)
kwargs_for_matrix = {}

# For compare_implementations, use the add/multiply implementations
compare_implementations_args = (
    {"add": add_numbers, "multiply": multiply_numbers},
)
compare_implementations_kwargs = {
    "args": args_for_add_multiply,
    "kwargs": kwargs_for_add_multiply,
    "warmup": 2,
    "iterations": 5,
}

# Build the final ARGS dict for compare_implementations
ARGS = {
    "implementations": {"add": add_numbers, "multiply": multiply_numbers},
    "args": (5, 3),
    "kwargs": {},
    "warmup": 2,
    "iterations": 5,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tensorflow — 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
compute_face_normals Ran
ran on a synthesized input
pmclSF/DeepCompress/src/ds_mesh_to_pc.py
code served (permissive licence) · get_code("ec68ad90a83f5021")
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 a simple mesh with 4 vertices and 2 triangular faces
vertices = np.array([
    [0.0, 0.0, 0.0],
    [1.0, 0.0, 0.0],
    [0.0, 1.0, 0.0],
    [0.0, 0.0, 1.0]
], dtype=np.float32)

# Define 2 triangular faces (each face is a triplet of vertex indices)
faces = np.array([
    [0, 1, 2],
    [0, 1, 3]
], dtype=np.int32)

ARGS = {
    "vertices": vertices,
    "faces": faces
}
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
count_points_in_block Ran
ran on a synthesized input
pmclSF/DeepCompress/src/ds_select_largest.py
code served (permissive licence) · get_code("d59a566b22bb35e7")
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 int, ndim 0, dtype family int, shape () 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 tempfile
import os

# Create a temporary PLY file for testing
temp_dir = tempfile.mkdtemp()
ply_file_path = os.path.join(temp_dir, "test_block.ply")

# Write a valid PLY file with a header and some points
with open(ply_file_path, 'w') as f:
    f.write("ply\n")
    f.write("format ascii 1.0\n")
    f.write("element vertex 5\n")
    f.write("property float x\n")
    f.write("property float y\n")
    f.write("property float z\n")
    f.write("end_header\n")
    f.write("0.0 0.0 0.0\n")
    f.write("1.0 1.0 1.0\n")
    f.write("2.0 2.0 2.0\n")
    f.write("3.0 3.0 3.0\n")
    f.write("4.0 4.0 4.0\n")

ARGS = {
    "file_path": ply_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 int, ndim 0, dtype family int, shape () — shape () 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
create_model Ran
ran on a synthesized input
pmclSF/DeepCompress/src/cli_train.py
code served (permissive licence) · get_code("f584cbf9232ff4a4")
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 Sequential. 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 tensorflow as tf

class HyperparameterConfig:
    """Stand-in for Keras Tuner hyperparameter object"""
    def __init__(self):
        self._values = {
            'num_layers': 2,
            'layer_0_units': 128,
            'layer_1_units': 256,
            'learning_rate': 0.001
        }
    
    def Int(self, name, min_value=None, max_value=None, step=None):
        """Return a fixed integer value for hyperparameter tuning"""
        if name in self._values:
            return self._values[name]
        # Return a default value within the specified range
        if min_value is not None:
            return min_value
        return 1
    
    def Float(self, name, min_value=None, max_value=None, sampling=None):
        """Return a fixed float value for hyperparameter tuning"""
        if name in self._values:
            return self._values[name]
        # Return a default value within the specified range
        if min_value is not None:
            return min_value
        return 0.001

hp = HyperparameterConfig()

ARGS = {
    'hp': hp
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Sequential — a Sequential output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Sequential 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
create_test_input Ran
ran on a synthesized input
pmclSF/DeepCompress/src/benchmarks.py
code served (permissive licence) · get_code("999c397d5e575b9f")
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 EagerTensor, ndim 5, dtype family float, shape (2, 4, 3, 5, 6) 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 tensorflow as tf

ARGS = {
    "batch_size": 2,
    "depth": 4,
    "height": 3,
    "width": 5,
    "channels": 6,
    "dtype": tf.float32,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type EagerTensor, ndim 5, dtype family float, shape (2, 4, 3, 5, 6) — shape (2, 4, 3, 5, 6) 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_colorbar Ran
ran on a synthesized input
pmclSF/DeepCompress/src/colorbar.py
code served (permissive licence) · get_code("077e192bc1b492c0")
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 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 matplotlib as mpl
import matplotlib.pyplot as plt
import numpy as np

ARGS = {
    'vmin': 0.0,
    'vmax': 1.0,
    'cmap': 'inferno',
    'orientation': 'horizontal',
    'figsize': (8, 1),
    'title': 'Test Colorbar',
    'tick_labels': ['Low', 'Medium', 'High'],
    'tick_positions': [0.0, 0.5, 1.0],
    'label_format': '{:.2f}',
    'font_size': 10,
    'title_size': 12,
    'tick_rotation': 45,
    'extend': 'neither'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
matplotlib, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, 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_log2_constant Ran
ran on a synthesized input
pmclSF/DeepCompress/src/constants.py
code served (permissive licence) · get_code("c55084ec2dcca697")
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 EagerTensor, ndim 0, dtype family float, shape () 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 math
import tensorflow as tf

LOG_2 = tf.constant(math.log(2.0), dtype=tf.float32, name='log_2')
LOG_2_F16 = tf.constant(math.log(2.0), dtype=tf.float16, name='log_2_f16')

def get_log2_constant(dtype=tf.float32):
    """Get LOG_2 constant in the specified dtype."""
    if dtype == tf.float16:
        return LOG_2_F16
    return LOG_2

ARGS = {
    'dtype': tf.float32
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type EagerTensor, ndim 0, dtype family float, shape () — shape () 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_log2_reciprocal Ran
ran on a synthesized input
pmclSF/DeepCompress/src/constants.py
code served (permissive licence) · get_code("24bc9d78e458a628")
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 SymbolicTensor. 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 tensorflow as tf

# Test case 1: default dtype (float32)
ARGS_1 = {
    'dtype': tf.float32
}

# Test case 2: float16 dtype
ARGS_2 = {
    'dtype': tf.float16
}

# Use the default case for ARGS
ARGS = ARGS_1
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type SymbolicTensor — a SymbolicTensor 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
load_and_preprocess_data Ran
ran on a synthesized input
pmclSF/DeepCompress/src/cli_train.py
code served (permissive licence) · get_code("29c977decc3bde78")
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 DatasetV1Adapter. 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 numpy as np
import tensorflow as tf

# Create a mock read_point_cloud function since it's called in the function
def read_point_cloud(file_path):
    """Mock function that returns a numpy array of vertices"""
    np.random.seed(0)
    return np.random.randn(2048, 3).astype(np.float32)

# Create temporary directory with mock .ply files
temp_dir = tempfile.mkdtemp()
for i in range(2):
    ply_file = os.path.join(temp_dir, f'cloud_{i}.ply')
    with open(ply_file, 'w') as f:
        f.write('dummy ply content')

ARGS = {
    'input_dir': temp_dir,
    'batch_size': 2
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type DatasetV1Adapter — a DatasetV1Adapter output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a DatasetV1Adapter 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
read_json Ran
ran on a synthesized input
mauriceqch/pcc_geo_cnn_v2/src/ev_compare.py
code served (permissive licence) · get_code("9cc30d373e78edd4")
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
import tempfile
import os

# Create a temporary JSON file
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, "test.json")

# Write a simple JSON object to the file
test_data = {
    "key1": "value1",
    "key2": 42,
    "key3": [1, 2, 3],
    "key4": {"nested": "object"}
}

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

ARGS = {
    "file": 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 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
read_off Ran
ran on a synthesized input
pmclSF/DeepCompress/src/ds_mesh_to_pc.py
code served (permissive licence) · get_code("c8c2a7859c98b84a")
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 MeshData. 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
import os
import numpy as np

# Create a temporary OFF file for testing
temp_dir = tempfile.mkdtemp()
off_file_path = os.path.join(temp_dir, "test_mesh.off")

# Write a simple OFF file with a tetrahedron
with open(off_file_path, 'w') as f:
    f.write("OFF\n")
    f.write("4 4 0\n")  # 4 vertices, 4 faces, 0 edges
    f.write("0.0 0.0 0.0\n")
    f.write("1.0 0.0 0.0\n")
    f.write("0.5 1.0 0.0\n")
    f.write("0.5 0.5 1.0\n")
    f.write("3 0 1 2\n")
    f.write("3 0 1 3\n")
    f.write("3 1 2 3\n")
    f.write("3 2 0 3\n")

ARGS = {
    "file_path": off_file_path
}
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 MeshData — a MeshData output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a MeshData 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
sample_points_from_mesh Ran
ran on a synthesized input
pmclSF/DeepCompress/src/ds_mesh_to_pc.py
code served (permissive licence) · get_code("2896d6b1fca2e0f7")
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 numpy as np
from dataclasses import dataclass
from typing import Optional

@dataclass
class MeshData:
    """Container for mesh data including vertices, faces, and normals."""
    vertices: np.ndarray
    faces: Optional[np.ndarray] = None
    vertex_normals: Optional[np.ndarray] = None
    face_normals: Optional[np.ndarray] = None

# Create a simple mesh with 4 vertices and 2 triangular faces
vertices = np.array([
    [0.0, 0.0, 0.0],
    [1.0, 0.0, 0.0],
    [0.0, 1.0, 0.0],
    [1.0, 1.0, 0.0]
], dtype=np.float32)

faces = np.array([
    [0, 1, 2],
    [1, 3, 2]
], dtype=np.int32)

# Compute face normals
v1s = vertices[faces[:, 0]]
v2s = vertices[faces[:, 1]]
v3s = vertices[faces[:, 2]]
face_normals = np.cross(v2s - v1s, v3s - v1s)
face_normals = face_normals / np.linalg.norm(face_normals, axis=1, keepdims=True)

# Compute vertex normals (simple average of adjacent face normals)
vertex_normals = np.array([
    [0.0, 0.0, 1.0],
    [0.0, 0.0, 1.0],
    [0.0, 0.0, 1.0],
    [0.0, 0.0, 1.0]
], dtype=np.float32)

mesh_data = MeshData(
    vertices=vertices,
    faces=faces,
    vertex_normals=vertex_normals,
    face_normals=face_normals
)

ARGS = {
    "mesh_data": mesh_data,
    "num_points": 256,
    "compute_normals": True
}
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, 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

Also present in other papers' code

Of 15 samples lifted from this paper's linked repositories, 15 carry a function identity and 0 do not. Of the 15 distinct functions with an identity, 1 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 1 the same month, 1 later) and 14 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
flatten 21 other papers
6 dated earlier, 1 the same month, 14 later
1609.08675 (2016-09)
YouTube-8M: A Large-Scale Video Classification Benchmark
1802.04821 (2018-02)
Evolved Policy Gradients
1811.05014 (2018-11)
NeXtVLAD: An Efficient Neural Network to Aggregate Frame-level Features for Large-scale Video Classification
and 18 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 2026-11-21.

Repositories linked to this paper

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

Abstract

Point clouds have been recognized as a crucial data structure for 3D content and are essential in a number of applications such as virtual and mixed reality, autonomous driving, cultural heritage, etc. In this paper, we propose a set of contributions to improve deep point cloud compression, i.e.: using a scale hyperprior model for entropy coding; employing deeper transforms; a different balancing weight in the focal loss; optimal thresholding for decoding; and sequential model training. In addition, we present an extensive ablation study on the impact of each of these factors, in order to provide a better understanding about why they improve RD performance. An optimal combination of the proposed improvements achieves BD-PSNR gains over G-PCC trisoup and octree of 5.50 (6.48) dB and 6.84 (5.95) dB, respectively, when using the point-to-point (point-to-plane) metric. Code is available at https://github.com/mauriceqch/pcc_geo_cnn_v2 .

For agents

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

get_harvested_code_for_paper("2006.09043")
get_code_for_paper("2006.09043")
have("2006.09043")

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

curl https://syntology.ai/api/ran/2006.09043.json

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

[![Syntology run record](https://syntology.ai/api/ran/2006.09043.svg)](https://syntology.ai/paper/2006.09043)

Connect an agent — have() is free.