SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2206.01724 · NeurIPS · 2022

SNAKE: Shape-aware Neural 3D Keypoint Field

Chuang Gan, Fuchun Sun, Wenbing Huang, Chengliang Zhong, Peixing You, Xiaoxue Chen, Hao Zhao, Guyue Zhou, Xiaodong Mu, Keypoint Coord, Saliency Prob, Coord Recon, and 2 more

arXiv · PDF · Open in the Atlas

Code that ran

11 of 14 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 4 have no contract verdict; 6 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/2206.01724.json · badge: /api/ran/2206.01724.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
zhongcl-thu/SNAKE canonical (classifier label)
classifier basis not recorded on the link
11 ran · 3 no recorded run
FunctionStatusWhere it lives
accuracy Ran
ran on a synthesized input
zhongcl-thu/SNAKE/core/utils/common.py
code served (permissive licence) · get_code("9a8db8d09e7923d3")
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, ndim 2, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# The function expects:
# - output: a tensor with topk method, shape (batch_size, num_classes)
# - target: a tensor of shape (batch_size,) with class indices
# - topk: tuple of k values (default (1,))
# - v_n: batch size used for normalization (default 256)

batch_size = 2
num_classes = 4

# Create output tensor - logits or scores
# Shape: (batch_size, num_classes)
output = torch.randn(batch_size, num_classes)

# Create target tensor - class indices
# Shape: (batch_size,)
target = torch.tensor([0, 2])

# topk parameter
topk = (1,)

# v_n parameter - batch size for normalization
v_n = 2

ARGS = {
    "output": output,
    "target": target,
    "topk": topk,
    "v_n": v_n
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 2, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
angles2rotation_matrix Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
zhongcl-thu/SNAKE/core/datasets/augmentation.py
pointer only (licence: MIT); the stored copy get_code serves, from rsy6318/CorrI2P, is pointer only (licence: NONE) · get_code("60f2f4207ae20ab9")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from rsy6318/CorrI2P, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

atomic_rotate_pytorch_batch Ran
ran; our contract draft was wrong, not the code
zhongcl-thu/SNAKE/core/datasets/augmentation.py
code served (permissive licence) · get_code("61969ba9c876c10d")
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 Tensor, ndim 3, dtype family float, shape (2, 3, 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 numpy as np
import torch

torch.manual_seed(0)

# Create a small batch of 3D atomic coordinates
# Shape: Bx3xN where B=batch size, 3=coordinates (x,y,z), N=number of atoms
B = 2  # batch size
N = 4  # number of atoms
data = torch.randn(B, 3, N, dtype=torch.float32)

# Create rotation angles in radians [theta_x, theta_y, theta_z]
angles = [0.1, 0.2, 0.3]

ARGS = {
    "data": data,
    "angles": angles,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 3, 4) — shape (2, 3, 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
augment Ran
ran on a synthesized input
zhongcl-thu/SNAKE/core/datasets/augmentation.py
code served (permissive licence) · get_code("60c37454e2a095ce")
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 small point cloud data (5 points, 3D coordinates)
pcd_data = np.random.randn(5, 3).astype(np.float32)

# Create grid coordinates (3 points, 3D coordinates)
grid_coords = np.random.randn(3, 3).astype(np.float32)

# Create config dictionary with all required parameters
config = {
    'do_aug': True,
    'rotate_angle': 45.0,
    'rotate_dim': '3D',
    'translation_perturbation': True,
    'shift_x': [-0.1, 0.1],
    'shift_y': [-0.1, 0.1],
    'shift_z': [-0.1, 0.1],
}

ARGS = {
    'pcd_data': pcd_data,
    'grid_coords': grid_coords,
    'config': config,
}
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
chamfer_distance_naive Ran
ran; our contract draft was wrong, not the code
zhongcl-thu/SNAKE/core/nets/common.py
code served (permissive licence) · get_code("8e6b5059f61d84e6")
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 Tensor, ndim 1, dtype family float, shape (2,) 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 torch
import numpy as np

torch.manual_seed(0)

# The function expects two tensors of point clouds
# Shape: (batch_size, num_points, 3)
# The function uses .size(), .view(), .pow(), .sum(), .min(), .mean()
# So we need torch tensors, not numpy arrays

batch_size = 2
num_points = 4
point_dim = 3

points1 = torch.randn(batch_size, num_points, point_dim)
points2 = torch.randn(batch_size, num_points, point_dim)

ARGS = {
    "points1": points1,
    "points2": points2,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 1, dtype family float, shape (2,) — shape (2,) 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
compute_iou Ran
ran, honoured the contract we drafted
zhongcl-thu/SNAKE/core/nets/common.py
code served (permissive licence) · get_code("0072de07b144b152")
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 1, dtype family float, shape (2,) 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 torch

torch.manual_seed(0)

# Create two small occupancy tensors
# occ1: shape (2, 4) - batch of 2, flattened occupancy values of 4
# occ2: shape (2, 4) - same shape for comparison
occ1 = torch.rand(2, 4)
occ2 = torch.rand(2, 4)

ARGS = {
    "occ1": occ1,
    "occ2": occ2,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (2,) — shape (2,) 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_ray_directions_cam Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
zhongcl-thu/SNAKE/core/nets/geometry.py
code served (permissive licence) · get_code("dcf31906f80d979c")
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 Tensor, ndim 3, dtype family float, shape (2, 4, 3) as a later run recorded it. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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 torch
import torch.nn as nn

torch.manual_seed(0)

# uv: Shape (batch_size, num_samples, 2)
# Contains pixel coordinates (x, y)
batch_size = 2
num_samples = 4
uv = torch.randn(batch_size, num_samples, 2)

# intrinsics: Shape (batch_size, 3, 3)
# Camera intrinsic matrix with fx, fy, cx, cy
intrinsics = torch.zeros(batch_size, 3, 3)
intrinsics[:, 0, 0] = 500.0  # fx
intrinsics[:, 1, 1] = 500.0  # fy
intrinsics[:, 0, 2] = 320.0  # cx
intrinsics[:, 1, 2] = 240.0  # cy

ARGS = {
    "uv": uv,
    "intrinsics": intrinsics,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 4, 3) — shape (2, 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
load_last_iter Ran
ran on a synthesized input
zhongcl-thu/SNAKE/core/utils/common.py
code served (permissive licence) · get_code("97c47b2462e29db3")
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 os
import torch
import tempfile

torch.manual_seed(0)

# Create a temporary checkpoint file with the expected structure
with tempfile.NamedTemporaryFile(mode='wb', delete=False, suffix='.pt') as f:
    checkpoint_path = f.name
    checkpoint = {
        'step': 42
    }
    torch.save(checkpoint, checkpoint_path)

ARGS = {
    'path': checkpoint_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — 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
make_3d_grid Ran
ran; our contract draft was wrong, not the code
zhongcl-thu/SNAKE/core/datasets/train_dataset.py
code served (permissive licence) · get_code("9babaac7293b5582")
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 (60, 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

# Analyze the function:
# - x_res, y_res, z_res are expected to be tuples/lists with 3 elements: [start, end, num_points]
# - type is a string that can be 'DHW' or 'HWD'

# Create small test ranges for the coordinate system
x_res = (0.0, 1.0, 5)  # 5 points from 0 to 1
y_res = (0.0, 2.0, 4)  # 4 points from 0 to 2
z_res = (0.0, 3.0, 3)  # 3 points from 0 to 3

ARGS = {
    'x_res': x_res,
    'y_res': y_res,
    'z_res': z_res,
    'type': 'DHW'
}
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 (60, 3) — shape (60, 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
reflect_vector_on_vector Ran
ran; our contract draft was wrong, not the code
zhongcl-thu/SNAKE/core/nets/geometry.py
code served (permissive licence) · get_code("54dc24b743feff79")
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 Tensor, ndim 2, dtype family float, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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 torch
import torch.nn.functional as F

torch.manual_seed(0)

# Create two small tensors for the reflection operation
# batch size: 2, dimension: 4
vector_to_reflect = torch.randn(2, 4)
reflection_axis = torch.randn(2, 4)

ARGS = {
    "vector_to_reflect": vector_to_reflect,
    "reflection_axis": reflection_axis,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, 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
sample_vertex_from_mesh Ran
ran on a synthesized input
zhongcl-thu/SNAKE/core/datasets/smpl_model.py
code served (permissive licence) · get_code("0bd31bdc88752b70")
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 small mesh data
# vertex: (num_vertices, 3) array of vertex positions
num_vertices = 6
vertex = np.random.rand(num_vertices, 3).astype(np.float32)

# facet: (num_triangles, 3) array of vertex indices forming triangles
num_triangles = 4
facet = np.array([
    [0, 1, 2],
    [1, 2, 3],
    [2, 3, 4],
    [3, 4, 5]
], dtype=np.int32)

# colors: (num_vertices, 3) array of RGB colors
colors = np.random.rand(num_vertices, 3).astype(np.float32)

# rnd_idxs: None or array of random triangle indices
rnd_idxs = None

# u: None or (num_samples, 1) barycentric coordinate
u = None

# v: None or (num_samples, 1) barycentric coordinate
v = None

# num_samples: number of points to sample
num_samples = 2048

ARGS = {
    "vertex": vertex,
    "facet": facet,
    "colors": colors,
    "rnd_idxs": rnd_idxs,
    "u": u,
    "v": v,
    "num_samples": num_samples
}
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
chamfer_distance Not yet run
unverified (harvested, no recorded run)
zhongcl-thu/SNAKE/core/nets/common.py
code served (permissive licence) · get_code("f4d8f4c4d0f7da71")
How it ran

no invocation recorded.

load_state Not yet run
unverified (harvested, no recorded run)
zhongcl-thu/SNAKE/core/utils/common.py
code served (permissive licence) · get_code("3777dfeb89e049b9")
How it ran

no invocation recorded.

naive_read_pcd Not yet run
unverified (harvested, no recorded run)
zhongcl-thu/SNAKE/core/datasets/train_dataset.py
code served (permissive licence) · get_code("c994f6f4b55acda6")
How it ran

no invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 carry a function identity and 0 do not. Of the 14 distinct functions with an identity, 9 are also present in other papers whose code we hold (6 with at least one paper dated earlier, 0 the same month, 9 later) and 5 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
compute_iou 9 other papers
5 dated earlier, 4 later
2003.04618 (2020-03)
Convolutional Occupancy Networks
2104.01148 (2021-04)
Decomposing 3D Scenes into Objects via Unsupervised Volume Segmentation
2107.01899 (2021-07)
Ray-ONet: Efficient 3D Reconstruction From A Single RGB Image
and 6 more not named here
chamfer_distance_naive 6 other papers
3 dated earlier, 3 later
2003.04618 (2020-03)
Convolutional Occupancy Networks
2107.01899 (2021-07)
Ray-ONet: Efficient 3D Reconstruction From A Single RGB Image
2204.01264 (2022-04)
Probabilistic Implicit Scene Completion
and 3 more not named here
chamfer_distance 3 other papers
2 dated earlier, 1 later
2107.01899 (2021-07)
Ray-ONet: Efficient 3D Reconstruction From A Single RGB Image
2204.01264 (2022-04)
Probabilistic Implicit Scene Completion
2309.05098 (2023-09)
3D Implicit Transporter for Temporally Consistent Keypoint Discovery
get_ray_directions_cam 3 other papers
1 dated earlier, 2 later
1906.01618 (2019-06)
Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representations
2211.09786 (2022-11)
SE(3)-Equivariant Relational Rearrangement with Neural Descriptor Fields
2309.05098 (2023-09)
3D Implicit Transporter for Temporally Consistent Keypoint Discovery
naive_read_pcd 3 other papers
2 dated earlier, 1 later
2002.12687 (2020-02)
KeypointNet: A Large-scale 3D Keypoint Dataset Aggregated from Numerous Human Annotations
2109.01605 (2021-09)
Representing Shape Collections with Alignment-Aware Linear Models
2308.11916 (2023-08)
Semantic-Aware Implicit Template Learning via Part Deformation Consistency
reflect_vector_on_vector 3 other papers
1 dated earlier, 2 later
1906.01618 (2019-06)
Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representations
2211.09786 (2022-11)
SE(3)-Equivariant Relational Rearrangement with Neural Descriptor Fields
2309.05098 (2023-09)
3D Implicit Transporter for Temporally Consistent Keypoint Discovery
angles2rotation_matrix 2 other papers
2 later
2309.05098 (2023-09)
3D Implicit Transporter for Temporally Consistent Keypoint Discovery
2401.12452 (2024-01)
Self-supervised Learning of LiDAR 3D Point Clouds via 2D-3D Neural Calibration
atomic_rotate_pytorch_batch 1 other paper
1 later
2309.05098 (2023-09)
3D Implicit Transporter for Temporally Consistent Keypoint Discovery
make_3d_grid 1 other paper
1 later
2309.05098 (2023-09)
3D Implicit Transporter for Temporally Consistent Keypoint Discovery

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

Detecting 3D keypoints from point clouds is important for shape reconstruction, while this work investigates the dual question: can shape reconstruction benefit 3D keypoint detection? Existing methods either seek salient features according to statistics of different orders or learn to predict keypoints that are invariant to transformation. Nevertheless, the idea of incorporating shape reconstruction into 3D keypoint detection is under-explored. We argue that this is restricted by former problem formulations. To this end, a novel unsupervised paradigm named SNAKE is proposed, which is short for shape-aware neural 3D keypoint field. Similar to recent coordinate-based radiance or distance field, our network takes 3D coordinates as inputs and predicts implicit shape indicators and keypoint saliency simultaneously, thus naturally entangling 3D keypoint detection and shape reconstruction. We achieve superior performance on various public benchmarks, including standalone object datasets ModelNet40, KeypointNet, SMPL meshes and scene-level datasets 3DMatch and Redwood. Intrinsic shape awareness brings several advantages as follows. (1) SNAKE generates 3D keypoints consistent with human semantic annotation, even without such supervision. (2) SNAKE outperforms counterparts in terms of repeatability, especially when the input point clouds are down-sampled. (3) the generated keypoints allow accurate geometric registration, notably in a zero-shot setting. Codes are available at https://github.com/zhongcl-thu/SNAKE.

For agents

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

get_harvested_code_for_paper("2206.01724")
get_code_for_paper("2206.01724")
have("2206.01724")

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

curl https://syntology.ai/api/ran/2206.01724.json

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

[![Syntology run record](https://syntology.ai/api/ran/2206.01724.svg)](https://syntology.ai/paper/2206.01724)

Connect an agent — have() is free.