SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1903.08548 · 2019

Learning Convolutional Transforms for Lossy Point Cloud Geometry Compression

arXiv · PDF · Open in the Atlas

Code that ran

10 of 13 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 10 have no contract verdict; 0 failures of our instrument; 3 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/1903.08548.json · badge: /api/ran/1903.08548.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 canonical (classifier label)
classifier basis: census
10 ran · 3 no recorded run
FunctionStatusWhere it lives
bdrate Ran
ran on a synthesized input
mauriceqch/pcc_geo_cnn/src/metrics.py
code served (permissive licence) · get_code("380e43980da3f3f4")
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 float, 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 numpy

# Examine the function: it expects metric_set1 and metric_set2
# Each metric_set should be iterable of tuples where:
# - First element (x[0]) is rate (used in math.log)
# - Second element (x[1]) is psnr (used in numpy.polyfit)
# The function fits a 3rd-degree polynomial, so we need at least 4 points

# Create two rate-distortion curves with realistic values
# Curve 1: lower bitrates, lower PSNR
metric_set1 = [
    (100, 28.5),
    (200, 31.2),
    (400, 33.8),
    (800, 35.6),
]

# Curve 2: higher bitrates, higher PSNR (better codec/settings)
metric_set2 = [
    (80, 29.2),
    (160, 32.1),
    (320, 34.9),
    (640, 36.8),
]

ARGS = {
    "metric_set1": metric_set1,
    "metric_set2": metric_set2,
}
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 float, 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
bdsnr Ran
ran on a synthesized input
mauriceqch/pcc_geo_cnn/src/metrics.py
code served (permissive licence) · get_code("4055858887d2f567")
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 float64, ndim 0, 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 math
import numpy

# The function expects two metric_sets
# Each metric_set is a list of tuples where each tuple is (rate, psnr)
# The function extracts rate and psnr, takes log of rates, fits polynomials,
# and computes the Bjontegaard metric

# Create metric_set1: rate-distortion curve 1
# rates should be positive values (since math.log is called on them)
metric_set1 = [
    (100.0, 30.5),
    (500.0, 35.2),
    (1000.0, 38.1),
    (2000.0, 40.3),
]

# Create metric_set2: rate-distortion curve 2
# Different curve with overlapping rate range
metric_set2 = [
    (150.0, 31.0),
    (600.0, 36.0),
    (1200.0, 39.0),
    (2500.0, 41.5),
]

ARGS = {
    "metric_set1": metric_set1,
    "metric_set2": metric_set2,
}
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 float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
compress Ran
ran on a synthesized input
mauriceqch/pcc_geo_cnn/src/compress.py
code served (permissive licence) · get_code("68ecd4d88b382314")
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 bytes. 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

# Construct valid arguments for the compress function
nn_output = {
    'x_shape': [4, 3],  # Small distinct dimensions
    'y_shape': [5, 2],  # Different dimensions to distinguish from x_shape
    'string': b'test_compressed_data'  # Bytes string for the compressed data
}

ARGS = {
    'nn_output': nn_output
}
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 bytes — a bytes output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a bytes 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
df_to_pc Ran
ran on a synthesized input
mauriceqch/pcc_geo_cnn/src/pc_io.py
code served (permissive licence) · get_code("247fff5ca5bf5dcd")
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 PC. 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
import pandas as pd

# Create a small dataframe with x, y, z columns
df = pd.DataFrame({
    'x': np.array([0.5, 1.5, 2.5]),
    'y': np.array([0.5, 1.5, 2.5]),
    'z': np.array([0.5, 1.5, 2.5])
})

# Define bounds that satisfy all assertions
p_min = np.array([0.0, 0.0, 0.0])
p_max = np.array([3.0, 3.0, 3.0])

ARGS = {
    'df': df,
    'p_min': p_min,
    'p_max': p_max
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type PC — a PC output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a PC 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
focal_loss Ran
ran on a synthesized input
mauriceqch/pcc_geo_cnn/src/focal_loss.py
code served (permissive licence) · get_code("427c81f065a9770a")
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 tensorflow as tf
import tensorflow.keras.backend as K
import numpy as np

# Create small test tensors for focal_loss
# y_true: binary labels (0 or 1)
# y_pred: predicted probabilities (values between 0 and 1)

batch_size = 2
num_classes = 1

# y_true: binary labels
y_true = tf.constant([[1.0], [0.0]], dtype=tf.float32)

# y_pred: predicted probabilities
y_pred = tf.constant([[0.9], [0.3]], dtype=tf.float32)

# gamma and alpha are hyperparameters for focal loss
gamma = 2
alpha = 0.95

ARGS = {
    "y_true": y_true,
    "y_pred": y_pred,
    "gamma": gamma,
    "alpha": alpha
}
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 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
load_compressed_file Ran
ran on a synthesized input
mauriceqch/pcc_geo_cnn/src/decompress.py
code served (permissive licence) · get_code("f86eb9601b8309ac")
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 numpy as np
import gzip
import tempfile
import os

# Create a temporary gzip file with valid compressed data
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, "test_compressed.gz")

# Write test data: x_shape (3 uint16 values), y_shape (3 uint16 values), and some payload
with gzip.open(temp_file, 'wb') as f:
    # Write x_shape: 3 uint16 values
    x_shape = np.array([2, 4, 8], dtype=np.uint16)
    f.write(x_shape.tobytes())
    
    # Write y_shape: 3 uint16 values
    y_shape = np.array([2, 3, 5], dtype=np.uint16)
    f.write(y_shape.tobytes())
    
    # Write some payload string
    f.write(b"test payload data")

ARGS = {
    "file": temp_file
}
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 — 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
pa_to_df Ran
ran on a synthesized input
mauriceqch/pcc_geo_cnn/src/pc_io.py
code served (permissive licence) · get_code("f1707746d8ee4270")
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 DataFrame. 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
import pandas as pd

# Create a small point array with shape (2, 3) - 2 points with x, y, z coordinates
points = np.array([
    [1.0, 2.0, 3.0],
    [4.0, 5.0, 6.0]
], dtype=np.float32)

ARGS = {
    'points': points
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type DataFrame — a DataFrame output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a DataFrame output the recorded ndim 2 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
pc_to_df Ran
ran on a synthesized input
mauriceqch/pcc_geo_cnn/src/pc_io.py
code served (permissive licence) · get_code("fcdb1e7181b8f30c")
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 DataFrame. 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
import torch

class MockPointCloud:
    def __init__(self, points):
        self.points = points

torch.manual_seed(0)

# Create a small point cloud with 4 points and 3 coordinates (x, y, z)
points_array = np.array([
    [1.0, 2.0, 3.0],
    [4.0, 5.0, 6.0],
    [7.0, 8.0, 9.0],
    [10.0, 11.0, 12.0]
], dtype=np.float32)

pc = MockPointCloud(points_array)

ARGS = {
    'pc': pc
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type DataFrame — a DataFrame output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a DataFrame output the recorded ndim 2 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
process_x Ran
ran on a synthesized input
mauriceqch/pcc_geo_cnn/src/compression_model.py
code served (permissive licence) · get_code("e7fc6f25be90ce85")
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 2, dtype family float, shape (2, 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 tensorflow as tf
import numpy as np

# Create a sparse tensor
# SparseTensor requires indices, values, and dense_shape
indices = np.array([[0, 0], [1, 2], [1, 3]], dtype=np.int64)
values = np.array([1.0, 2.0, 3.0], dtype=np.float32)
dense_shape = np.array([2, 4], dtype=np.int64)

x = tf.SparseTensor(indices=indices, values=values, dense_shape=dense_shape)

# Define the dense_tensor_shape that matches what we expect after conversion
dense_tensor_shape = (2, 4)

ARGS = {
    "x": x,
    "dense_tensor_shape": dense_tensor_shape,
}
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 EagerTensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 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
quantize_tensor Ran
ran on a synthesized input
mauriceqch/pcc_geo_cnn/src/compression_model.py
code served (permissive licence) · get_code("a3383ecf338746d7")
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 2, dtype family int, shape (2, 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 tensorflow as tf
import numpy as np

# Create a small TensorFlow tensor with values in range [0, 1]
x = tf.constant([[0.2, 0.5, 0.8, 1.0],
                  [0.0, 0.3, 0.7, 0.9]], dtype=tf.float32)

ARGS = {
    'x': x
}
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 EagerTensor, ndim 2, dtype family int, shape (2, 4) — shape (2, 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
input_fn Not yet run
unverified (harvested, no recorded run)
mauriceqch/pcc_geo_cnn/src/decompress.py
code served (permissive licence) · get_code("b3538a22f65deaa2")
How it ran

no invocation recorded.

load_compressed_files Not yet run
unverified (harvested, no recorded run)
mauriceqch/pcc_geo_cnn/src/decompress.py
code served (permissive licence) · get_code("750bc0b5e83d1873")
How it ran

no invocation recorded.

pc_to_tf Not yet run
unverified (harvested, no recorded run)
mauriceqch/pcc_geo_cnn/src/compression_model.py
code served (permissive licence) · get_code("606deb2380489bba")
How it ran

no invocation recorded.

Also present in other papers' code

Of 13 samples lifted from this paper's linked repositories, 13 carry a function identity and 0 do not. Of the 13 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 13 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

Efficient point cloud compression is fundamental to enable the deployment of virtual and mixed reality applications, since the number of points to code can range in the order of millions. In this paper, we present a novel data-driven geometry compression method for static point clouds based on learned convolutional transforms and uniform quantization. We perform joint optimization of both rate and distortion using a trade-off parameter. In addition, we cast the decoding process as a binary classification of the point cloud occupancy map. Our method outperforms the MPEG reference solution in terms of rate-distortion on the Microsoft Voxelized Upper Bodies dataset with 51.5% BDBR savings on average. Moreover, while octree-based methods face exponential diminution of the number of points at low bitrates, our method still produces high resolution outputs even at low bitrates. Code and supplementary material are available at https://github.com/mauriceqch/pcc_geo_cnn .

For agents

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

get_harvested_code_for_paper("1903.08548")
get_code_for_paper("1903.08548")
have("1903.08548")

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

curl https://syntology.ai/api/ran/1903.08548.json

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

[![Syntology run record](https://syntology.ai/api/ran/1903.08548.svg)](https://syntology.ai/paper/1903.08548)

Connect an agent — have() is free.