SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2402.16086 · AAAI · 2024

Deep Homography Estimation for Visual Place Recognition

Lijun Zhang, Chun Yuan, Dongmei Jiang, Xiangyuan Lan, Feng Lu, Shuting Dong, Bingxi Liu

arXiv · PDF · Open in the Atlas

Code that ran

10 of 17 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; 4 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 6 failures of our instrument; 7 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/2402.16086.json · badge: /api/ran/2402.16086.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
lu-feng/dhe-vpr canonical (classifier label)
classifier basis: census
9 ran · 6 no recorded run
Lu-Feng/DHE-VPR community
Papers with Code archive
1 ran · 1 no recorded run
FunctionStatusWhere it lives
collate_fn Ran
ran; our contract draft was wrong, not the code
lu-feng/dhe-vpr/datasets_ws.py
code served (permissive licence) · get_code("937b98b0302097b0")
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 torch

torch.manual_seed(0)

# Create a single batch element
# images: (12, 3, h, w) - 12 images with 3 channels
images = torch.randn(12, 3, 4, 4)

# triplets_local_indexes: (10, 3) - 10 triplets with 3 indices each
triplets_local_indexes = torch.randint(0, 12, (10, 3))

# triplets_global_indexes: (12,) - 12 global indices
triplets_global_indexes = torch.arange(12)

# Create batch with 2 elements
batch_elem_1 = (
    images.clone(),
    triplets_local_indexes.clone(),
    triplets_global_indexes.clone()
)

batch_elem_2 = (
    images.clone(),
    triplets_local_indexes.clone(),
    triplets_global_indexes.clone()
)

batch = [batch_elem_1, batch_elem_2]

ARGS = {
    "batch": batch
}
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 — 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
compute_similarity Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Lu-Feng/DHE-VPR/network.py
code served (permissive licence) · get_code("daf11df89eb6cd54")
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 4, dtype family float, shape (2, 15, 3, 5) 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 torch.nn.functional as F

torch.manual_seed(0)

# Create small feature tensors with distinct dimensions
# Shape: (batch, channels, height, width)
batch_size = 2
channels = 4
height = 3
width = 5

features_a = torch.randn(batch_size, channels, height, width)
features_b = torch.randn(batch_size, channels, height, width)

ARGS = {
    "features_a": features_a,
    "features_b": features_b,
}
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 4, dtype family float, shape (2, 15, 3, 5) — shape (2, 15, 3, 5) 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
gem Ran
ran; our contract draft was wrong, not the code
lu-feng/dhe-vpr/network.py
code served (permissive licence) · get_code("ad695e1e64bd704c")
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 4, dtype family float, shape (2, 4, 1, 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 torch
import torch.nn.functional as F

torch.manual_seed(0)

# The gem function expects:
# - x: a 4D tensor (batch, channels, height, width) for avg_pool2d
# - p: a numeric parameter (default 3)
# - eps: a small epsilon value (default 1e-06)

# Create a small 4D tensor: batch=2, channels=4, height=3, width=5
x = torch.randn(2, 4, 3, 5)

# p is the power parameter
p = 3

# eps is the epsilon for clamping
eps = 1e-06

ARGS = {
    "x": x,
    "p": p,
    "eps": eps
}
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 4, dtype family float, shape (2, 4, 1, 1) — shape (2, 4, 1, 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
get_keypoints Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
lu-feng/dhe-vpr/homography_project.py
code served (permissive licence) · get_code("4db942a401f79a6d")
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 int, shape (384, 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

# The function get_keypoints takes img_size as a parameter
# img_size is unpacked as H, W (height and width)
# It should be a tuple or list of two integers

img_size = (256, 384)  # H=256, W=384 (distinct dimensions)

ARGS = {
    "img_size": img_size
}
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 int, shape (384, 2) — shape (384, 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_output_dim Ran
ran, honoured the contract we drafted
lu-feng/dhe-vpr/commons.py
code served (permissive licence) · get_code("a45a1ebcd76f102e")
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 torch
import torch.nn as nn

torch.manual_seed(0)

class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(3 * 224 * 224, 128)
    
    def forward(self, x):
        batch_size = x.shape[0]
        x = x.view(batch_size, -1)
        x = self.linear(x)
        x = x.view(batch_size, 128, 1, 1)
        return x

model = SimpleModel()

ARGS = {
    'model': model,
    'pooling_type': 'gem'
}
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
match_batch_tensor Ran
ran; our contract draft was wrong, not the code
lu-feng/dhe-vpr/homography_project.py
code served (permissive licence) · get_code("dd744fafe75ef469")
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 0, dtype family float, shape () 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 numpy as np

torch.manual_seed(0)

# Build arguments for match_batch_tensor

# fm1: (l, D) - feature map 1
l = 5  # number of keypoints
D = 4  # feature dimension
fm1 = torch.randn(l, D)

# fm2: (N, l, D) - batch of feature maps
N = 2  # batch size
fm2 = torch.randn(N, l, D)

# thetas: (N, 3, 3) - homography matrices for batch
thetas = torch.randn(N, 3, 3)

# trainflag: boolean
trainflag = True

# img_size: (H, W) - image height and width
# Must be divisible by patch_size (16)
img_size = (64, 80)

ARGS = {
    "fm1": fm1,
    "fm2": fm2,
    "thetas": thetas,
    "trainflag": trainflag,
    "img_size": img_size
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, 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 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
open_image_and_apply_transform Ran
ran, honoured the contract we drafted
lu-feng/dhe-vpr/datasets_util.py
code served (permissive licence) · get_code("3af50595c400a1d0")
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 (3, 384, 384) 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
from PIL import Image
import io

# Create a small test image in memory
# Create a simple RGB image (100x100 pixels)
img_array = np.random.randint(0, 256, (100, 100, 3), dtype=np.uint8)
pil_img = Image.fromarray(img_array, mode='RGB')

# Save to a temporary path-like string using BytesIO
# We'll create a temporary file-like object, but since we can't do real file I/O,
# we need to work around this. Instead, we'll mock the function's behavior by
# creating an image file and passing its path through a workaround.

# Create a simple PNG in memory and save it temporarily
import tempfile
import os

# Create a temporary directory and file
temp_dir = tempfile.gettempdir()
temp_image_path = os.path.join(temp_dir, "test_image_temp.png")

# Save the test image
pil_img.save(temp_image_path)

# Define ARGS with the temporary image path
ARGS = {
    "image_path": temp_image_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy, torch, torchvision — 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 (3, 384, 384) — shape (3, 384, 384) 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
open_image_and_apply_transform_train Ran
ran, honoured the contract we drafted
lu-feng/dhe-vpr/datasets_util.py
code served (permissive licence) · get_code("c91f6ee832367388")
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 (3, 384, 384) 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
from PIL import Image
import tempfile
import os

# Create a temporary directory and a small test image
temp_dir = tempfile.mkdtemp()
image_path = os.path.join(temp_dir, "test_image.png")

# Create a small RGB image (e.g., 100x100 pixels)
test_image = Image.new('RGB', (100, 100), color=(73, 109, 137))
test_image.save(image_path)

ARGS = {
    "image_path": image_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy, torch, torchvision — 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 (3, 384, 384) — shape (3, 384, 384) 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
path_to_pil_img Ran
ran, honoured the contract we drafted
lu-feng/dhe-vpr/datasets_ws.py
code served (permissive licence) · get_code("f0db6d80ebf12bfa")
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 Image. 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
from PIL import Image
import io

# Create a minimal valid image as bytes in memory
# Create a small 4x3 RGB image
img = Image.new('RGB', (4, 3), color=(255, 0, 0))
img_bytes = io.BytesIO()
img.save(img_bytes, format='PNG')
img_bytes.seek(0)

# Since the function uses Image.open(path), we need to provide a path
# We'll create a temporary file-like object, but we need an actual path string
# For testing purposes, we'll use a data URI approach or mock

# Actually, let's create a simple approach: use a string path
# Since we can't do actual file I/O, we'll need to mock or use a workaround
# The simplest is to create a temporary file

import tempfile
import os

# Create a temporary directory and file
temp_dir = tempfile.gettempdir()
temp_path = os.path.join(temp_dir, 'test_image.png')

# Create and save a test image to a real temporary file
test_img = Image.new('RGB', (4, 3), color=(255, 0, 0))
test_img.save(temp_path)

ARGS = {
    'path': temp_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Image — a Image output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Image output the recorded ndim 3 and dtype family int 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
to_cuda Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
lu-feng/dhe-vpr/train_dhe.py
code served (permissive licence) · get_code("e0db474b009dabf3")
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. 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)

# Create a list of tensors and modules that have .cuda() methods
tensor1 = torch.randn(2, 4)
tensor2 = torch.randn(3, 5)

class SimpleModule(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)

module1 = SimpleModule()

# Build the list_ parameter with items that have .cuda() method
list_ = [tensor1, tensor2, module1]

ARGS = {
    "list_": list_
}
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 — a list 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
Network Not yet run
unverified (harvested, no recorded run)
Lu-Feng/DHE-VPR/network.py
code served (permissive licence) · get_code("35a2d5e5ea929ea3")
How it ran

no invocation recorded.

cct_2 Not yet run
unverified (harvested, no recorded run)
lu-feng/dhe-vpr/model/cct/cct.py
code served (permissive licence) · get_code("cf97fc0d1f953674")
How it ran

no invocation recorded.

cct_4 Not yet run
unverified (harvested, no recorded run)
lu-feng/dhe-vpr/model/cct/cct.py
code served (permissive licence) · get_code("f709bdf67ed60e8a")
How it ran

no invocation recorded.

cct_6 Not yet run
unverified (harvested, no recorded run)
lu-feng/dhe-vpr/model/cct/cct.py
code served (permissive licence) · get_code("f1b8c16700889467")
How it ran

no invocation recorded.

compute_features Not yet run
unverified (harvested, no recorded run)
lu-feng/dhe-vpr/util.py
code served (permissive licence) · get_code("b3145a842e257fe6")
How it ran

no invocation recorded.

compute_recalls Not yet run
unverified (harvested, no recorded run)
lu-feng/dhe-vpr/util.py
code served (permissive licence) · get_code("73fee8a3f2a6bee4")
How it ran

no invocation recorded.

reprojection_error_ofinliers Not yet run
unverified (harvested, no recorded run)
lu-feng/dhe-vpr/homography_project.py
code served (permissive licence) · get_code("2f631cc89e3a62e9")
How it ran

no invocation recorded.

Also present in other papers' code

Of 17 samples lifted from this paper's linked repositories, 17 carry a function identity and 0 do not. Of the 17 distinct functions with an identity, 6 are also present in other papers whose code we hold (6 with at least one paper dated earlier, 2 the same month, 3 later) and 11 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
path_to_pil_img 9 other papers
3 dated earlier, 1 the same month, 5 later
2204.03444 (2022-04)
Deep Visual Geo-localization Benchmark
2306.02994 (2023-06)
Long-range UAV Thermal Geo-localization with Satellite Imagery
2310.05184 (2023-10)
AANet: Aggregation and Alignment Network with Semi-hard Positive Sample Mining for Hierarchical Place Recognition
and 6 more not named here
collate_fn 7 other papers
3 dated earlier, 1 the same month, 3 later
2204.03444 (2022-04)
Deep Visual Geo-localization Benchmark
2306.02994 (2023-06)
Long-range UAV Thermal Geo-localization with Satellite Imagery
2310.05184 (2023-10)
AANet: Aggregation and Alignment Network with Semi-hard Positive Sample Mining for Hierarchical Place Recognition
and 4 more not named here
gem 7 other papers
6 dated earlier, 1 later
1711.02512 (2017-11)
Fine-tuning CNN Image Retrieval with No Human Annotation
1901.10332 (2019-01)
Who's Afraid of Adversarial Queries? The Impact of Image Modifications on Content-based Image Retrieval
2001.08972 (2020-01)
SOLAR: Second-Order Loss and Attention for Image Retrieval
and 4 more not named here
cct_2 2 other papers
2 dated earlier
2204.03444 (2022-04)
Deep Visual Geo-localization Benchmark
2310.05184 (2023-10)
AANet: Aggregation and Alignment Network with Semi-hard Positive Sample Mining for Hierarchical Place Recognition
cct_4 2 other papers
2 dated earlier
2204.03444 (2022-04)
Deep Visual Geo-localization Benchmark
2310.05184 (2023-10)
AANet: Aggregation and Alignment Network with Semi-hard Positive Sample Mining for Hierarchical Place Recognition
cct_6 2 other papers
2 dated earlier
2204.03444 (2022-04)
Deep Visual Geo-localization Benchmark
2310.05184 (2023-10)
AANet: Aggregation and Alignment Network with Semi-hard Positive Sample Mining for Hierarchical Place Recognition

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

Visual place recognition (VPR) is a fundamental task for many applications such as robot localization and augmented reality. Recently, the hierarchical VPR methods have received considerable attention due to the trade-off between accuracy and efficiency. They usually first use global features to retrieve the candidate images, then verify the spatial consistency of matched local features for re-ranking. However, the latter typically relies on the RANSAC algorithm for fitting homography, which is time-consuming and nondifferentiable. This makes existing methods compromise to train the network only in global feature extraction. Here, we propose a transformer-based deep homography estimation (DHE) network that takes the dense feature map extracted by a backbone network as input and fits homography for fast and learnable geometric verification. Moreover, we design a reprojection error of inliers loss to train the DHE network without additional homography labels, which can also be jointly trained with the backbone network to help it extract the features that are more suitable for local matching. Extensive experiments on benchmark datasets show that our method can outperform several state-of-the-art methods. And it is more than one order of magnitude faster than the mainstream hierarchical VPR methods using RANSAC. The code is released at https://github.com/Lu-Feng/DHE-VPR.

For agents

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

get_harvested_code_for_paper("2402.16086")
get_code_for_paper("2402.16086")
have("2402.16086")

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

curl https://syntology.ai/api/ran/2402.16086.json

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

[![Syntology run record](https://syntology.ai/api/ran/2402.16086.svg)](https://syntology.ai/paper/2402.16086)

Connect an agent — have() is free.