SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1810.09381 · 2018

Unsupervised Learning of Shape and Pose with Differentiable Point Clouds

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/1810.09381.json · badge: /api/ran/1810.09381.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
eldar/differentiable-point-clouds community (classifier label)
classifier basis not recorded on the link
10 ran · 3 no recorded run
FunctionStatusWhere it lives
drc_loss Ran
ran on a synthesized input
eldar/differentiable-point-clouds/dpc/util/losses.py
code served (permissive licence) · get_code("60ee93b1c66fab31")
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 numpy as np

class Config:
    def __init__(self):
        self.vox_size = 2

cfg = Config()

# Create small tensors for the function
# probs shape: [vox_size+1, batch, height, width, depth] = [3, 2, 4, 5, 3]
probs = tf.constant(
    np.random.rand(3, 2, 4, 5, 3).astype(np.float32),
    dtype=tf.float32
)

# gt_proj shape: [batch, height, width, depth] = [2, 4, 5, 3]
gt_proj = tf.constant(
    np.random.rand(2, 4, 5, 3).astype(np.float32),
    dtype=tf.float32
)

ARGS = {
    "cfg": cfg,
    "probs": probs,
    "gt_proj": gt_proj
}
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
get_dropout_prob Ran
ran on a synthesized input
eldar/differentiable-point-clouds/dpc/models/model_pc.py
code served (permissive licence) · get_code("8df602630b1c437a")
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

class Config:
    def __init__(self):
        self.pc_point_dropout_scheduled = True
        self.pc_point_dropout = 0.1
        self.pc_point_dropout_exponential_schedule = False
        self.max_number_of_steps = 1000
        self.pc_point_dropout_start_step = 100
        self.pc_point_dropout_end_step = 900

cfg = Config()
global_step = tf.constant(500, dtype=tf.int32)

ARGS = {
    "cfg": cfg,
    "global_step": global_step
}
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_smooth_sigma Ran
ran on a synthesized input
eldar/differentiable-point-clouds/dpc/models/model_pc.py
code served (permissive licence) · get_code("6e6dd4087fa03641")
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; 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 tensorflow as tf

class Config:
    def __init__(self):
        self.max_number_of_steps = 100
        self.pc_relative_sigma_end = 0.5
        self.pc_relative_sigma = 0.1

cfg = Config()
global_step = 25

ARGS = {
    "cfg": cfg,
    "global_step": global_step
}
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 — no output shape is stated: observe_signature.py could not observe it (import:TypeError)
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
multi_expand Ran
ran on a synthesized input
eldar/differentiable-point-clouds/dpc/util/point_cloud.py
code served (permissive licence) · get_code("647d04e2eb2ab580")
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 (1, 1, 1, 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 test tensor
inp = tf.constant(np.random.randn(2, 4), dtype=tf.float32)

# Define axis and num parameters
axis = 0
num = 3

ARGS = {
    'inp': inp,
    'axis': axis,
    'num': num
}
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 5, dtype family float, shape (1, 1, 1, 2, 4) — shape (1, 1, 1, 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
obj_centened_camera_pos Ran
ran on a synthesized input
eldar/differentiable-point-clouds/dpc/render/render_point_cloud_blender.py
code served (permissive licence) · get_code("7ce659b80c0f9983")
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 math

def obj_centened_camera_pos(dist, azimuth_deg, elevation_deg):
    phi = float(elevation_deg) / 180 * math.pi
    theta = float(azimuth_deg) / 180 * math.pi
    x = dist * math.cos(theta) * math.cos(phi)
    y = dist * math.sin(theta) * math.cos(phi)
    z = dist * math.sin(phi)
    return (x, y, z)

ARGS = {
    "dist": 5.0,
    "azimuth_deg": 45.0,
    "elevation_deg": 30.0
}
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 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
pointcloud2voxels Ran
ran on a synthesized input
eldar/differentiable-point-clouds/dpc/util/point_cloud.py
code served (permissive licence) · get_code("1813da43b573bc7b")
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, 4, 4, 1) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import tensorflow as tf

class Config:
    def __init__(self):
        self.vox_size = 4
        self.pc_normalise_gauss = False
        self.pc_normalise_gauss_analytical = False

cfg = Config()

batch_size = 2
num_points = 5
input_pc = tf.random.normal((batch_size, num_points, 3))

sigma = 0.5

ARGS = {
    'cfg': cfg,
    'input_pc': input_pc,
    'sigma': sigma
}
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, 4, 4, 1) — shape (2, 4, 4, 4, 1) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
pointcloud2voxels3d_fast Ran
ran on a synthesized input
eldar/differentiable-point-clouds/dpc/util/point_cloud.py
code served (permissive licence) · get_code("720152235094b576")
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 tensorflow as tf

class Config:
    def __init__(self):
        self.vox_size = 4
        self.vox_size_z = 5
        self.pc_rgb_stop_points_gradient = False

cfg = Config()

batch_size = 2
num_points = 6

pc = tf.random.uniform([batch_size, num_points, 3], minval=-0.5, maxval=0.5, dtype=tf.float32)

rgb = tf.random.uniform([batch_size, num_points, 3], minval=0.0, maxval=1.0, dtype=tf.float32)

ARGS = {
    "cfg": cfg,
    "pc": pc,
    "rgb": rgb
}
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 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
pool_single_view Ran
ran on a synthesized input
eldar/differentiable-point-clouds/dpc/models/model_base.py
code served (permissive licence) · get_code("dd16754ef3053ae3")
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, 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 tensorflow as tf

class Config:
    def __init__(self, batch_size, step_size):
        self.batch_size = batch_size
        self.step_size = step_size

cfg = Config(batch_size=2, step_size=4)

tensor = tf.constant([
    [1.0, 2.0, 3.0],
    [4.0, 5.0, 6.0],
    [7.0, 8.0, 9.0],
    [10.0, 11.0, 12.0],
    [13.0, 14.0, 15.0],
    [16.0, 17.0, 18.0],
    [19.0, 20.0, 21.0],
    [22.0, 23.0, 24.0],
])

view_idx = 1

ARGS = {
    "cfg": cfg,
    "tensor": tensor,
    "view_idx": view_idx,
}
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 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
tf_repeat_0 Ran
ran on a synthesized input
eldar/differentiable-point-clouds/dpc/models/model_pc.py
code served (permissive licence) · get_code("5614bddbbf44e931")
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 3, dtype family float, shape (10, 4, 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 tensorflow as tf
import numpy as np

# Create a small input tensor for testing
# Using shape (2, 4, 3) - batch size 2, 4 features, 3 dimensions
input_tensor = tf.constant(np.random.randn(2, 4, 3), dtype=tf.float32)

# Set num to repeat along the expanded dimension
num_repeats = 5

ARGS = {
    "input": input_tensor,
    "num": num_repeats
}
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 3, dtype family float, shape (10, 4, 3) — shape (10, 4, 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
voxel2pc Ran
ran on a synthesized input
eldar/differentiable-point-clouds/dpc/render/render_point_cloud_blender.py
code served (permissive licence) · get_code("d3d5b0f87687609f")
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

# Create a small voxel grid
# The function expects voxels to be squeezable to a 3D array
# Create a 4x4x4 voxel grid with some values above and below threshold
voxel_data = np.random.rand(1, 4, 4, 4) * 2.0  # Values between 0 and 2

# Threshold value to separate occupied from empty voxels
threshold_value = 0.5

ARGS = {
    "voxels": voxel_data,
    "threshold": threshold_value
}
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
drc_rgb_loss Not yet run
unverified (harvested, no recorded run)
eldar/differentiable-point-clouds/dpc/util/losses.py
code served (permissive licence) · get_code("c8f563881aa98dd4")
How it ran

no invocation recorded.

point_cloud_distance Not yet run
unverified (harvested, no recorded run)
eldar/differentiable-point-clouds/dpc/util/point_cloud_distance.py
code served (permissive licence) · get_code("6181cff74f8613bd")
How it ran

no invocation recorded.

regularization_loss Not yet run
unverified (harvested, no recorded run)
eldar/differentiable-point-clouds/dpc/util/losses.py
code served (permissive licence) · get_code("eecdb72ca340d002")
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, 7 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 7 later) and 6 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
drc_loss 1 other paper
1 later
2108.03746 (2021-08)
Unsupervised Learning of Fine Structure Generation for 3D Point Clouds by 2D Projection Matching
drc_rgb_loss 1 other paper
1 later
2108.03746 (2021-08)
Unsupervised Learning of Fine Structure Generation for 3D Point Clouds by 2D Projection Matching
get_dropout_prob 1 other paper
1 later
2108.03746 (2021-08)
Unsupervised Learning of Fine Structure Generation for 3D Point Clouds by 2D Projection Matching
get_smooth_sigma 1 other paper
1 later
2108.03746 (2021-08)
Unsupervised Learning of Fine Structure Generation for 3D Point Clouds by 2D Projection Matching
pool_single_view 1 other paper
1 later
2108.03746 (2021-08)
Unsupervised Learning of Fine Structure Generation for 3D Point Clouds by 2D Projection Matching
regularization_loss 1 other paper
1 later
2108.03746 (2021-08)
Unsupervised Learning of Fine Structure Generation for 3D Point Clouds by 2D Projection Matching
tf_repeat_0 1 other paper
1 later
2108.03746 (2021-08)
Unsupervised Learning of Fine Structure Generation for 3D Point Clouds by 2D Projection Matching

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

We address the problem of learning accurate 3D shape and camera pose from a collection of unlabeled category-specific images. We train a convolutional network to predict both the shape and the pose from a single image by minimizing the reprojection error: given several views of an object, the projections of the predicted shapes to the predicted camera poses should match the provided views. To deal with pose ambiguity, we introduce an ensemble of pose predictors which we then distill to a single "student" model. To allow for efficient learning of high-fidelity shapes, we represent the shapes by point clouds and devise a formulation allowing for differentiable projection of these. Our experiments show that the distilled ensemble of pose predictors learns to estimate the pose accurately, while the point cloud representation allows to predict detailed shape models. The supplementary video can be found at https://www.youtube.com/watch?v=LuIGovKeo60

For agents

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

get_harvested_code_for_paper("1810.09381")
get_code_for_paper("1810.09381")
have("1810.09381")

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

curl https://syntology.ai/api/ran/1810.09381.json

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

[![Syntology run record](https://syntology.ai/api/ran/1810.09381.svg)](https://syntology.ai/paper/1810.09381)

Connect an agent — have() is free.