SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2204.02287 · CVPR · 2022

Rethinking Visual Geo-localization for Large-Scale Applications

Barbara Caputo, Gabriele Berton, Carlo Masone

arXiv · PDF · Open in the Atlas

Code that ran

6 of 10 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (4 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 5 have no contract verdict; 1 failure of our instrument; 4 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2204.02287.json · badge: /api/ran/2204.02287.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
gmberton/CosPlace community
Papers with Code archive
4 ran · 3 no recorded run
stschubert/vpr_tutorial community
Papers with Code archive
1 ran · 1 no recorded run
yxgeee/OpenIBL community
Papers with Code archive
1 ran · 0 no recorded run
FunctionStatusWhere it lives
Flatten Ran
ran on a synthesized input
gmberton/CosPlace/cosplace_model/cosplace_network.py
code served (permissive licence) · get_code("bcef70112e3e16de")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, 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

INIT_ARGS = {}

ARGS = {
    "x": torch.randn(2, 4, 1, 1)
}
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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 — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
GeM Ran
ran on a synthesized input · constructed an object; no result was computed
gmberton/CosPlace/cosplace_model/cosplace_network.py
code served (permissive licence) · get_code("05b61e284a5179a3")
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 an instance of its own class, GeM: the run constructed the object and did not call it, so no result was computed. 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 torch import nn
import torch.nn.functional as F
from torch.nn.parameter import Parameter

torch.manual_seed(0)

# GeM expects:
# - p: learnable parameter (default 3)
# - eps: small float for numerical stability (default 1e-06)
# No other parameters in __init__

# For testing forward pass, we need an input tensor x
# The forward method uses x.clamp, .pow, F.avg_pool2d with kernel (x.size(-2), x.size(-1))
# So x needs to be 4D (batch, channels, height, width)

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

ARGS = {
    'p': 3,
    'eps': 1e-06
}
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 GeM — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a GeM 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
ImageDataset Ran
ran on a synthesized input · constructed an object; no result was computed
stschubert/vpr_tutorial/feature_extraction/feature_extractor_cosplace.py
pointer only (licence: GPL-3.0) · get_code("d0085f34014c954b")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, stschubert/vpr_tutorial, is not in the permissive set Syntology serves inline. 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.

L2Norm Ran
ran on a synthesized input · constructed an object; no result was computed
gmberton/CosPlace/cosplace_model/cosplace_network.py
code served (permissive licence) · get_code("a52c26b55a0e2d80")
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 an instance of its own class, L2Norm: the run constructed the object and did not call it, so no result was computed. 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)

# L2Norm.__init__ takes dim as a parameter
# L2Norm.forward takes x (a tensor) and applies F.normalize with p=2.0 and the specified dim

# Construct a small tensor for testing
x = torch.randn(2, 4)

ARGS = {
    'dim': 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 L2Norm — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a L2Norm 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
NetVLAD Ran
ran on a synthesized input · constructed an object; no result was computed
yxgeee/OpenIBL/ibl/models/netvlad.py
code served (permissive licence) · get_code("0ddf7bafa24d2ba9")
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 an instance of its own class, NetVLAD: the run constructed the object and did not call it, so no result was computed. 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
import numpy as np

torch.manual_seed(0)
np.random.seed(0)

ARGS = {
    'num_clusters': 8,
    'dim': 16,
    'alpha': 100.0,
    'normalize_input': True,
}
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 NetVLAD — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a NetVLAD 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
gem Ran
ran; our contract draft was wrong, not the code
gmberton/CosPlace/cosplace_model/cosplace_network.py
code served (permissive licence) · get_code("79f68aea8273d062")
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 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)

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

# Create a p parameter tensor (can be scalar or small tensor)
p = torch.ones(1) * 3

# eps is a float parameter
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, 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 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
CosPlaceFeatureExtractor Not yet run
unverified (harvested, no recorded run)
stschubert/vpr_tutorial/feature_extraction/feature_extractor_cosplace.py
pointer only (licence: GPL-3.0) · get_code("afaa2c01250f7753")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, stschubert/vpr_tutorial, is not in the permissive set Syntology serves inline. 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. No invocation recorded.

GeoLocalizationNet Not yet run
unverified (harvested, no recorded run)
gmberton/CosPlace/cosplace_model/cosplace_network.py
code served (permissive licence) · get_code("c726b15ccc929238")
How it ran

no invocation recorded.

get_backbone Not yet run
unverified (harvested, no recorded run)
gmberton/CosPlace/cosplace_model/cosplace_network.py
code served (permissive licence) · get_code("1522a0277165c6be")
How it ran

no invocation recorded.

get_pretrained_torchvision_model Not yet run
unverified (harvested, no recorded run)
gmberton/CosPlace/cosplace_model/cosplace_network.py
code served (permissive licence) · get_code("4102a46eb0cc1ac7")
How it ran

no invocation recorded.

Also present in other papers' code

Of 10 samples lifted from this paper's linked repositories, 10 carry a function identity and 0 do not. Of the 10 distinct functions with an identity, 6 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 6 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
gem
brick of this paper's assembly: helper
2 other papers
2 later
2308.10832 (2023-08)
EigenPlaces: Training Viewpoint Robust Models for Visual Place Recognition
2412.09199 (2024-12)
MutualVPR: A Mutual Learning Framework for Resolving Supervision Inconsistencies via Adaptive Clustering
Flatten
brick of this paper's assembly: helper
1 other paper
1 later · 1 only as an assembly brick
2308.10832 (2023-08)
EigenPlaces: Training Viewpoint Robust Models for Visual Place Recognition
GeM
brick of this paper's assembly: model
1 other paper
1 later · 1 only as an assembly brick
2308.10832 (2023-08)
EigenPlaces: Training Viewpoint Robust Models for Visual Place Recognition
ImageDataset
brick of this paper's assembly: helper
1 other paper
1 later · 1 only as an assembly brick
2308.10832 (2023-08)
EigenPlaces: Training Viewpoint Robust Models for Visual Place Recognition
L2Norm
brick of this paper's assembly: model
1 other paper
1 later · 1 only as an assembly brick
2308.10832 (2023-08)
EigenPlaces: Training Viewpoint Robust Models for Visual Place Recognition
get_pretrained_torchvision_model
brick of this paper's assembly: helper
1 other paper
1 later
2412.09199 (2024-12)
MutualVPR: A Mutual Learning Framework for Resolving Supervision Inconsistencies via Adaptive Clustering

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 Geo-localization (VG) is the task of estimating the position where a given photo was taken by comparing it with a large database of images of known locations. To investigate how existing techniques would perform on a real-world city-wide VG application, we build San Francisco eXtra Large, a new dataset covering a whole city and providing a wide range of challenging cases, with a size 30x bigger than the previous largest dataset for visual geo-localization. We find that current methods fail to scale to such large datasets, therefore we design a new highly scalable training technique, called CosPlace, which casts the training as a classification problem avoiding the expensive mining needed by the commonly used contrastive learning. We achieve state-of-the-art performance on a wide range of datasets and find that CosPlace is robust to heavy domain changes. Moreover, we show that, compared to the previous state-of-the-art, CosPlace requires roughly 80% less GPU memory at train time, and it achieves better results with 8x smaller descriptors, paving the way for city-wide real-world visual geo-localization. Dataset, code and trained models are available for research purposes at https://github.com/gmberton/CosPlace.

For agents

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

get_harvested_code_for_paper("2204.02287")
get_code_for_paper("2204.02287")
have("2204.02287")

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

curl https://syntology.ai/api/ran/2204.02287.json

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

[![Syntology run record](https://syntology.ai/api/ran/2204.02287.svg)](https://syntology.ai/paper/2204.02287)

Connect an agent — have() is free.