SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2312.03053 · 2023

Adaptive Multi-step Refinement Network for Robust Point Cloud Registration

arXiv · PDF · Open in the Atlas

Code that ran

4 of 4 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 0 have no contract verdict; 4 failures of our instrument; 0 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/2312.03053.json · badge: /api/ran/2312.03053.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
zhichen902/amr canonical (classifier label)
classifier basis: census
4 ran · 0 no recorded run
FunctionStatusWhere it lives
get_soft_mask Ran
ran; our contract draft was wrong, not the code
zhichen902/amr/experiments/geotransformer.3dmatch.stage4.gse.k3.max.oacl.stage2.sinkhorn/model.py
code served (permissive licence) · get_code("17d584be4ed0edfd")
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 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

torch.manual_seed(0)

# Small dimensions for k-nearest neighbors and point clouds
M = 2  # number of reference nodes
N = 2  # number of source nodes
K = 3  # number of nearest neighbors per node
D = 3  # spatial dimension (xyz coordinates)
num_ref_points = 4  # number of reference points
num_src_points = 5  # number of source points

# ref_knn_points: (M, K, 3) - k nearest neighbors for each reference node
ref_knn_points = torch.randn(M, K, D)

# src_knn_points: (N, K, 3) - k nearest neighbors for each source node
src_knn_points = torch.randn(N, K, D)

# transform: (4, 4) - homogeneous transformation matrix
transform = torch.eye(4)
transform[:3, :3] = torch.randn(3, 3)
transform[:3, 3] = torch.randn(3)

# pos_radius: float - positive correspondence radius
pos_radius = 0.5

# ref_points_f: (num_ref_points, 3) - reference frame points
ref_points_f = torch.randn(num_ref_points, D)

# src_points_f: (num_src_points, 3) - source frame points
src_points_f = torch.randn(num_src_points, D)

# ref_knn_masks: (M, K) - boolean mask for valid reference neighbors
ref_knn_masks = torch.ones(M, K, dtype=torch.bool)

# src_knn_masks: (N, K) - boolean mask for valid source neighbors
src_knn_masks = torch.ones(N, K, dtype=torch.bool)

ARGS = {
    "ref_knn_points": ref_knn_points,
    "src_knn_points": src_knn_points,
    "transform": transform,
    "pos_radius": pos_radius,
    "ref_points_f": ref_points_f,
    "src_points_f": src_points_f,
    "ref_knn_masks": ref_knn_masks,
    "src_knn_masks": src_knn_masks,
}
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 tuple, ndim 2, 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
inject_default_parser Ran
ran; our contract draft was wrong, not the code
zhichen902/amr/geotransformer/engine/base_tester.py
code served (permissive licence) · get_code("7e4df6976748d024")
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 ArgumentParser. 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 argparse

torch_manual_seed_result = None

ARGS = {
    "parser": None,
}
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 ArgumentParser — a ArgumentParser output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ArgumentParser 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
inject_default_parser Ran
ran; our contract draft was wrong, not the code
zhichen902/amr/geotransformer/engine/base_trainer.py
code served (permissive licence) · get_code("e59b3354578c7f3c")
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 ArgumentParser. 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 argparse

# The function inject_default_parser takes an optional parser argument
# and returns a parser with arguments added to it.
# Since it doesn't use the parser in complex ways (just adds arguments),
# we can pass either None or an existing ArgumentParser instance.

# Create a parser instance to pass
custom_parser = argparse.ArgumentParser(description="Test parser")

# Define ARGS mapping all parameters
ARGS = {
    'parser': custom_parser
}
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 ArgumentParser — a ArgumentParser output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ArgumentParser 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
prepare_one_data Ran
ran; our contract draft was wrong, not the code
zhichen902/amr/geotransformer/engine/epoch_based_trainer.py
code served (permissive licence) · get_code("a05a14e6dbbb0375")
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. 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 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)

# Build a minimal diffusion model that has the required methods
class MinimalDiffusion(nn.Module):
    def __init__(self):
        super().__init__()
        self.register_buffer('alphas', torch.linspace(0.1, 0.99, 100))
    
    def mix_prior_with_GT(self, gt_transform, estimated_transform, time_step):
        # time_step shape: (batch,)
        batch_size = gt_transform.shape[0]
        time_idx = time_step[0].long().item() % len(self.alphas)
        alpha = self.alphas[time_idx]
        sqrt_alpha_hat = torch.sqrt(alpha)
        sqrt_one_minus_alpha_hat = torch.sqrt(1 - alpha)
        
        # Mix transforms
        noised = sqrt_alpha_hat * gt_transform + sqrt_one_minus_alpha_hat * estimated_transform
        
        return noised, sqrt_alpha_hat, sqrt_one_minus_alpha_hat
    
    def mix_prior_with_GT_Slerp(self, gt_transform, estimated_transform, time_step):
        # Similar to mix_prior_with_GT but using Slerp
        batch_size = gt_transform.shape[0]
        time_idx = time_step[0].long().item() % len(self.alphas)
        alpha = self.alphas[time_idx]
        sqrt_alpha_hat = torch.sqrt(alpha)
        sqrt_one_minus_alpha_hat = torch.sqrt(1 - alpha)
        
        # Slerp-like interpolation
        noised = sqrt_alpha_hat * gt_transform + sqrt_one_minus_alpha_hat * estimated_transform
        
        return noised, sqrt_alpha_hat, sqrt_one_minus_alpha_hat

# Build a minimal engine_args config
class EngineArgs:
    def __init__(self, slerp=False):
        self.Slerp = slerp

# Setup
time_step = torch.tensor([5], dtype=torch.long)
diffusion = MinimalDiffusion()
engine_args = EngineArgs(slerp=False)

data_dict = {
    'transform': torch.randn(2, 4),
    'estimated_transform': torch.randn(2, 4),
}

ARGS = {
    'time_step': time_step,
    'data_dict': data_dict,
    'diffusion': diffusion,
    'engine_args': engine_args,
}
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 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

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 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 4 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

Point Cloud Registration (PCR) estimates the relative rigid transformation between two point clouds of the same scene. Despite significant progress with learning-based approaches, existing methods still face challenges when the overlapping region between the two point clouds is small. In this paper, we propose an adaptive multi-step refinement network that refines the registration quality at each step by leveraging the information from the preceding step. To achieve this, we introduce a training procedure and a refinement network. Firstly, to adapt the network to the current step, we utilize a generalized one-way attention mechanism, which prioritizes the last step's estimated overlapping region, and we condition the network on step indices. Secondly, instead of training the network to map either random transformations or a fixed pre-trained model's estimations to the ground truth, we train it on transformations with varying registration qualities, ranging from accurate to inaccurate, thereby enhancing the network's adaptiveness and robustness. Despite its conceptual simplicity, our method achieves state-of-the-art performance on both the 3DMatch/3DLoMatch and KITTI benchmarks. Notably, on 3DLoMatch, our method reaches 80.4% recall rate, with an absolute improvement of 1.2%.

For agents

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

get_harvested_code_for_paper("2312.03053")
get_code_for_paper("2312.03053")
have("2312.03053")

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

curl https://syntology.ai/api/ran/2312.03053.json

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

[![Syntology run record](https://syntology.ai/api/ran/2312.03053.svg)](https://syntology.ai/paper/2312.03053)

Connect an agent — have() is free.