SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · aaai_28123 · AAAI · 2024

VSFormer: Visual-Spatial Fusion Transformer for Correspondence Pruning

Li Zhao, Tao Wang, Xiaoqin Zhang, Guobao Xiao, Tangfei Liao, Vsfusion Contextformer

Code that ran

11 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, 1 violated it and 6 have no contract verdict; 4 failures of our instrument; 2 no recorded run), measured 2026-11-22; 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/aaai_28123.json · badge: /api/ran/aaai_28123.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
sugar-fly/VSFormer canonical (classifier label)
classifier basis: census
11 ran · 2 no recorded run
FunctionStatusWhere it lives
batch_episym Ran
ran; our contract draft was wrong, not the code
sugar-fly/VSFormer/models/loss.py
code served (permissive licence) · get_code("db1e63d4387aa9b7")
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. 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

torch.manual_seed(0)

# Analyze the function to understand what's needed:
# - x1: shape (batch_size, num_pts, 2) - 2D points that get converted to homogeneous (3,)
# - x2: shape (batch_size, num_pts, 2) - 2D points that get converted to homogeneous (3,)
# - F: Fundamental matrix, shape (..., 3, 3) that gets reshaped to (batch_size, 1, 3, 3)

batch_size = 2
num_pts = 4

# x1 and x2 are batches of 2D point correspondences
x1 = torch.randn(batch_size, num_pts, 2)
x2 = torch.randn(batch_size, num_pts, 2)

# F is a fundamental matrix (or batch of them)
# The function reshapes it with reshape(-1, 1, 3, 3), so it can be flat or (batch_size, 3, 3)
F = torch.randn(batch_size, 3, 3)

ARGS = {
    "x1": x1,
    "x2": x2,
    "F": F,
}
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
build_act_layer Ran
ran on a synthesized input
sugar-fly/VSFormer/models/vanilla_transformer.py
code served (permissive licence) · get_code("3680936f314fd5ee")
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 Identity. 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 torch
import torch.nn as nn
from typing import Union, Dict, Optional

torch.manual_seed(0)

# Test case 1: None input (should return Identity)
ARGS_1 = {
    'act_cfg': None
}

# Test case 2: String input (simple activation name)
ARGS_2 = {
    'act_cfg': 'ReLU'
}

# Test case 3: Dict input with activation type only
ARGS_3 = {
    'act_cfg': {'type': 'Tanh'}
}

# Test case 4: Dict input with LeakyReLU and custom negative_slope
ARGS_4 = {
    'act_cfg': {'type': 'LeakyReLU', 'negative_slope': 0.1}
}

# Test case 5: Dict input with LeakyReLU without negative_slope (should use default 0.2)
ARGS_5 = {
    'act_cfg': {'type': 'LeakyReLU'}
}

# Test case 6: Dict input with GELU
ARGS_6 = {
    'act_cfg': {'type': 'GELU'}
}

# Use the first test case as the primary ARGS
ARGS = ARGS_1
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 Identity — a Identity output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Identity 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
build_dropout_layer Ran
ran on a synthesized input
sugar-fly/VSFormer/models/vanilla_transformer.py
code served (permissive licence) · get_code("9f1963bba517e764")
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 Identity. 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 torch
import torch.nn as nn

torch.manual_seed(0)

# The function build_dropout_layer takes:
# - p: Optional[float] - dropout probability
# - **kwargs: additional arguments to pass to nn.Dropout

# Test case 1: p is None (should return nn.Identity)
ARGS = {
    "p": None,
}
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 Identity — a Identity output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Identity 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
collate_fn Ran
ran on a synthesized input
sugar-fly/VSFormer/datasets/CorresDataset.py
code served (permissive licence) · get_code("890be1553092cbaa")
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. 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

torch.manual_seed(0)

# Create sample data matching the structure expected by collate_fn
sample1 = {
    'K1': np.array([[500, 0, 320], [0, 500, 240], [0, 0, 1]], dtype=np.float32),
    'K2': np.array([[500, 0, 320], [0, 500, 240], [0, 0, 1]], dtype=np.float32),
    'T1': np.eye(4, dtype=np.float32),
    'T2': np.eye(4, dtype=np.float32),
    'R': np.eye(3, dtype=np.float32),
    't': np.array([0.1, 0.2, 0.3], dtype=np.float32),
    'virtPt': np.array([0, 0, 0], dtype=np.float32),
    'img1': np.random.rand(480, 640, 3).astype(np.float32),
    'img2': np.random.rand(480, 640, 3).astype(np.float32),
    'img1_path': '/path/to/img1.jpg',
    'img2_path': '/path/to/img2.jpg',
    'xs': np.random.rand(2, 8, 2).astype(np.float32),  # (2 images, 8 keypoints, 2 coords)
    'ys': np.random.randint(0, 2, 8).astype(np.float32),
    'side': np.random.rand(8, 4).astype(np.float32),
}

sample2 = {
    'K1': np.array([[500, 0, 320], [0, 500, 240], [0, 0, 1]], dtype=np.float32),
    'K2': np.array([[500, 0, 320], [0, 500, 240], [0, 0, 1]], dtype=np.float32),
    'T1': np.eye(4, dtype=np.float32),
    'T2': np.eye(4, dtype=np.float32),
    'R': np.eye(3, dtype=np.float32),
    't': np.array([0.15, 0.25, 0.35], dtype=np.float32),
    'virtPt': np.array([0.1, 0.1, 0.1], dtype=np.float32),
    'img1': np.random.rand(480, 640, 3).astype(np.float32),
    'img2': np.random.rand(480, 640, 3).astype(np.float32),
    'img1_path': '/path/to/img3.jpg',
    'img2_path': '/path/to/img4.jpg',
    'xs': np.random.rand(2, 6, 2).astype(np.float32),  # (2 images, 6 keypoints, 2 coords)
    'ys': np.random.randint(0, 2, 6).astype(np.float32),
    'side': np.random.rand(6, 4).astype(np.float32),
}

batch = [sample1, sample2]

ARGS = {
    'batch': batch
}
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 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
computeNN Ran
ran on a synthesized input
sugar-fly/VSFormer/dump_match/feature_match.py
code served (permissive licence) · get_code("11ecacc97cc9c582")
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. 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 numpy as np

torch.manual_seed(0)

# Create small descriptor arrays
# desc_ii: shape (2, 4) - 2 descriptors of dimension 4
# desc_jj: shape (3, 4) - 3 descriptors of dimension 4
desc_ii = np.random.randn(2, 4).astype(np.float32)
desc_jj = np.random.randn(3, 4).astype(np.float32)

ARGS = {
    'desc_ii': desc_ii,
    'desc_jj': desc_jj,
}
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 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
conv1x1 Ran
ran; our contract draft was wrong, not the code
sugar-fly/VSFormer/models/resnet34.py
code served (permissive licence) · get_code("d9def42110729a85")
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) and returned Conv2d. 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 torch
import torch.nn as nn

torch.manual_seed(0)

# The function conv1x1 takes in_planes, out_planes, and stride as parameters
# It returns an nn.Conv2d layer configured for 1x1 convolution
# We need to provide valid integer arguments for these parameters

ARGS = {
    "in_planes": 3,
    "out_planes": 64,
    "stride": 1
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Conv2d — a Conv2d output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Conv2d 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
conv3x3 Ran
ran; our contract draft was wrong, not the code
sugar-fly/VSFormer/models/resnet34.py
code served (permissive licence) · get_code("160bb14bd76201b4")
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) and returned Conv2d. 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 torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    "in_planes": 3,
    "out_planes": 64,
    "stride": 1,
    "groups": 1,
    "dilation": 1,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Conv2d — a Conv2d output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Conv2d 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
knn Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
sugar-fly/VSFormer/models/vsformer.py
code served (permissive licence) · get_code("5f563da12002f5b8")
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) and returned Tensor, ndim 3, dtype family int, shape (2, 3, 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

torch.manual_seed(0)

# x is used with:
# - transpose(2, 1): needs 3 dimensions
# - sum(x ** 2, dim=1, keepdim=True): sum over dimension 1
# - matmul(x.transpose(2, 1), x): matrix multiplication
# Shape should be (batch, features, num_points) based on the operations
# Let's use (2, 4, 3) - batch=2, features=4, num_points=3

x = torch.randn(2, 4, 3)

# k is the number of nearest neighbors to find
k = 2

ARGS = {
    "x": x,
    "k": k,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 int, shape (2, 3, 2) — shape (2, 3, 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
parse_cfg Ran
ran on a synthesized input
sugar-fly/VSFormer/models/vanilla_transformer.py
code served (permissive licence) · get_code("135c1dcf77712d97")
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 torch
from typing import Union, Dict, Optional, Tuple

torch.manual_seed(0)

# Test case 1: cfg as a string
cfg_string = "LinearLayer"

# Test case 2: cfg as a dict with 'type' key
cfg_dict = {
    'type': 'ConvLayer',
    'in_channels': 3,
    'out_channels': 64,
    'kernel_size': 3
}

ARGS = {
    'cfg': cfg_string
}
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 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
resnet34 Ran
ran on a synthesized input
sugar-fly/VSFormer/models/resnet34.py
code served (permissive licence) · get_code("c34df55bf3ffc13d")
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 ResNet. 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 torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    "in_channels": 3,
}
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 ResNet — a ResNet output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ResNet 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
str2bool Ran
ran, violated the contract we drafted
sugar-fly/VSFormer/dump_match/extract_feature.py
pointer only (licence: MIT); the stored copy get_code serves, from haoran0301/amenunet, is pointer only (licence: NONE) · get_code("248284f69adfeaad")
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 haoran0301/amenunet, 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.

batch_symeig Not yet run
unverified (harvested, no recorded run)
sugar-fly/VSFormer/models/vsformer.py
code served (permissive licence) · get_code("a6def402c1b8803b")
How it ran

no invocation recorded.

weighted_8points Not yet run
unverified (harvested, no recorded run)
sugar-fly/VSFormer/models/vsformer.py
code served (permissive licence) · get_code("c0a523711c071fd0")
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, 9 are also present in other papers whose code we hold (7 with at least one paper dated earlier, 0 the same month, 4 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.

FunctionAlso present inEarliest-datedOthers named
conv1x1 431 other papers
313 dated earlier, 30 later, 87 same year, order not determinable, 1 undated · 38 only as an assembly brick
1409.1556 (2014-09)
Very Deep Convolutional Networks for Large-Scale Image Recognition
1506.01186 (2015-06)
Cyclical Learning Rates for Training Neural Networks
1507.05717 (2015-07)
An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognition
and 428 more not named here
conv3x3 312 other papers
214 dated earlier, 22 later, 75 same year, order not determinable, 1 undated · 31 only as an assembly brick
1409.1556 (2014-09)
Very Deep Convolutional Networks for Large-Scale Image Recognition
1507.05717 (2015-07)
An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognition
1512.03385 (2015-12)
Deep Residual Learning for Image Recognition
and 309 more not named here
str2bool 33 other papers
29 dated earlier, 4 same year, order not determinable
1406.6247 (2014-06)
Recurrent Models of Visual Attention
1506.03134 (2015-06)
Pointer Networks
1611.09940 (2016-11)
Neural Combinatorial Optimization with Reinforcement Learning
and 30 more not named here
batch_episym 4 other papers
3 dated earlier, 1 same year, order not determinable
Zhong_T-Net_Effective_Permutation-Equivariant_Network_for_Two-View_Correspondence_Learning_ICCV_2021_paper (2021 (year only))
T-Net: Effective Permutation-Equivariant Network for Two-View Correspondence Learning
2101.00591 (2021-01)
Progressive Correspondence Pruning by Consensus Learning
aaai_25456 (2023 (year only))
ConvMatch: Rethinking Network Design for Two-View Correspondence Learning
and 1 more not named here
batch_symeig 2 other papers
2 dated earlier
Zhong_T-Net_Effective_Permutation-Equivariant_Network_for_Two-View_Correspondence_Learning_ICCV_2021_paper (2021 (year only))
T-Net: Effective Permutation-Equivariant Network for Two-View Correspondence Learning
aaai_25456 (2023 (year only))
ConvMatch: Rethinking Network Design for Two-View Correspondence Learning
computeNN 2 other papers
1 dated earlier, 1 same year, order not determinable
aaai_25456 (2023 (year only))
ConvMatch: Rethinking Network Design for Two-View Correspondence Learning
Zhang_DeMatch_Deep_Decomposition_of_Motion_Field_for_Two-View_Correspondence_Learning_CVPR_2024_paper (2024 (year only))
DeMatch: Deep Decomposition of Motion Field for Two-View Correspondence Learning
knn 2 other papers
2 dated earlier
2101.00591 (2021-01)
Progressive Correspondence Pruning by Consensus Learning
2202.06688 (2022-02)
Geometric Transformer for Fast and Robust Point Cloud Registration
build_dropout_layer 1 other paper
1 later
Liu_UNOPose_Unseen_Object_Pose_Estimation_with_an_Unposed_RGB-D_Reference_CVPR_2025_paper (2025 (year only))
UNOPose: Unseen Object Pose Estimation with an Unposed RGB-D Reference Image
parse_cfg 1 other paper
1 later
Liu_UNOPose_Unseen_Object_Pose_Estimation_with_an_Unposed_RGB-D_Reference_CVPR_2025_paper (2025 (year only))
UNOPose: Unseen Object Pose Estimation with an Unposed RGB-D Reference Image

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

Correspondence pruning aims to find correct matches (inliers) from an initial set of putative correspondences, which is a fundamental task for many applications. The process of finding is challenging, given the varying inlier ratios between scenes/image pairs due to significant visual differences. However, the performance of the existing methods is usually limited by the problem of lacking visual cues (e.g., texture, illumination, structure) of scenes. In this paper, we propose a Visual-Spatial Fusion Transformer (VSFormer) to identify inliers and recover camera poses accurately. Firstly, we obtain highly abstract visual cues of a scene with the cross attention between local features of two-view images. Then, we model these visual cues and correspondences by a joint visual-spatial fusion module, simultaneously embedding visual cues into correspondences for pruning. Additionally, to mine the consistency of correspondences, we also design a novel module that combines the KNN-based graph and the transformer, effectively capturing both local and global contexts. Extensive experiments have demonstrated that the proposed VSFormer outperforms state-of-the-art methods on outdoor and indoor benchmarks. Our code is provided at the following repository: https://github.com/sugar-fly/VSFormer.

For agents

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

get_harvested_code_for_paper("aaai_28123")
get_code_for_paper("aaai_28123")
have("aaai_28123")

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

curl https://syntology.ai/api/ran/aaai_28123.json

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

[![Syntology run record](https://syntology.ai/api/ran/aaai_28123.svg)](https://syntology.ai/paper/aaai_28123)

Connect an agent — have() is free.