SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2004.02693 · 2020

Finding Your (3D) Center: 3D Object Detection Using a Learned Loss

arXiv · PDF · Open in the Atlas

Code that ran

4 of 8 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 4 have no contract verdict; 0 failures of our instrument; 4 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/2004.02693.json · badge: /api/ran/2004.02693.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
dgriffiths3/finding-your-center community (classifier label)
classifier basis not recorded on the link
4 ran · 4 no recorded run
FunctionStatusWhere it lives
chamfer_distance Ran
ran on a synthesized input
dgriffiths3/finding-your-center/utils/losses.py
code served (permissive licence) · get_code("08fd7e4daf7b69ef")
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 numpy as np
import tensorflow as tf

# Set random seed for reproducibility
tf.random.set_seed(0)
np.random.seed(0)

# Create small test tensors
batch_size = 2
num_pred_points = 4
num_gt_points = 3
feature_dim = 3

# labels: batch_size x max_gt_points x 4 (x, y, z, mask)
# We need enough points so that when masked, we get at least some valid points
labels = np.zeros((batch_size, 5, 4), dtype=np.float32)
# Batch 0: 3 valid points with mask=1, 2 invalid points with mask=0
labels[0, :3, :3] = np.random.randn(3, 3)
labels[0, :3, 3] = 1  # valid mask
labels[0, 3:, 3] = 0  # invalid mask

# Batch 1: 3 valid points with mask=1, 2 invalid points with mask=0
labels[1, :3, :3] = np.random.randn(3, 3)
labels[1, :3, 3] = 1  # valid mask
labels[1, 3:, 3] = 0  # invalid mask

# pred: batch_size x num_pred_points x feature_dim
pred = np.random.randn(batch_size, num_pred_points, feature_dim).astype(np.float32)

ARGS = {
    "labels": labels,
    "pred": pred,
}
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
mean_squared_error Ran
ran on a synthesized input
dgriffiths3/finding-your-center/utils/losses.py
code served (permissive licence) · get_code("90544be04b46200a")
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

# Create small tensors for labels and predictions
# Using batch_size=2, feature_dim=4
labels = tf.constant([[1.0, 2.0, 3.0, 4.0], [5.0, 6.0, 7.0, 8.0]], dtype=tf.float32)
pred = tf.constant([[1.1, 2.1, 3.1, 4.1], [5.1, 6.1, 7.1, 8.1]], dtype=tf.float32)

ARGS = {
    "labels": labels,
    "pred": pred,
}
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
nn_distance Ran
ran on a synthesized input
dgriffiths3/finding-your-center/utils/losses.py
code served (permissive licence) · get_code("3b864f5cb86ee7b7")
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
import numpy as np

# Create small test tensors with distinct dimensions
# a: batch_size=2, n=4, feature_dim=3
# b: batch_size=2, m=5, feature_dim=3
a = tf.constant(np.random.randn(2, 4, 3).astype(np.float32))
b = tf.constant(np.random.randn(2, 5, 3).astype(np.float32))

ARGS = {
    "a": a,
    "b": b,
}
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 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
sample_and_group_all Ran
ran on a synthesized input
dgriffiths3/finding-your-center/pointnet2/pointnet_util.py
code served (permissive licence) · get_code("49345dd0c75de3e0")
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 tensorflow as tf

# Set seeds for reproducibility
np.random.seed(0)
tf.random.set_seed(0)

# Create small test tensors
batch_size = 2
nsample = 4
point_feature_dim = 3

# xyz: shape (batch_size, nsample, 3)
xyz = tf.constant(np.random.randn(batch_size, nsample, 3).astype(np.float32), dtype=tf.float32)

# points: shape (batch_size, nsample, point_feature_dim) - additional features per point
points = tf.constant(np.random.randn(batch_size, nsample, point_feature_dim).astype(np.float32), dtype=tf.float32)

# use_xyz: boolean flag to include xyz in concatenation
use_xyz = True

ARGS = {
    'xyz': xyz,
    'points': points,
    'use_xyz': use_xyz
}
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 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
forward_pass Not yet run
unverified (harvested, no recorded run)
dgriffiths3/finding-your-center/loss_network/evaluate.py
code served (permissive licence) · get_code("16b5531fc4acfa6a")
How it ran

no invocation recorded.

sample_and_group Not yet run
unverified (harvested, no recorded run)
dgriffiths3/finding-your-center/pointnet2/pointnet_util.py
code served (permissive licence) · get_code("815e2af0a954aa24")
How it ran

no invocation recorded.

update_test_metrics Not yet run
unverified (harvested, no recorded run)
dgriffiths3/finding-your-center/scene_network/metrics_util.py
code served (permissive licence) · get_code("90f1ab8cee0c976e")
How it ran

no invocation recorded.

update_train_metrics Not yet run
unverified (harvested, no recorded run)
dgriffiths3/finding-your-center/scene_network/metrics_util.py
code served (permissive licence) · get_code("5e5941338e00b419")
How it ran

no invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 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 8 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

Massive semantically labeled datasets are readily available for 2D images, however, are much harder to achieve for 3D scenes. Objects in 3D repositories like ShapeNet are labeled, but regrettably only in isolation, so without context. 3D scenes can be acquired by range scanners on city-level scale, but much fewer with semantic labels. Addressing this disparity, we introduce a new optimization procedure, which allows training for 3D detection with raw 3D scans while using as little as 5% of the object labels and still achieve comparable performance. Our optimization uses two networks. A scene network maps an entire 3D scene to a set of 3D object centers. As we assume the scene not to be labeled by centers, no classic loss, such as Chamfer can be used to train it. Instead, we use another network to emulate the loss. This loss network is trained on a small labeled subset and maps a non centered 3D object in the presence of distractions to its own center. This function is very similar - and hence can be used instead of - the gradient the supervised loss would provide. Our evaluation documents competitive fidelity at a much lower level of supervision, respectively higher quality at comparable supervision. Supplementary material can be found at: https://dgriffiths3.github.io.

For agents

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

get_harvested_code_for_paper("2004.02693")
get_code_for_paper("2004.02693")
have("2004.02693")

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

curl https://syntology.ai/api/ran/2004.02693.json

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

[![Syntology run record](https://syntology.ai/api/ran/2004.02693.svg)](https://syntology.ai/paper/2004.02693)

Connect an agent — have() is free.