Pith. sign in

REVIEW 5 major objections 5 minor 70 references

EDTformer: An Efficient Decoder Transformer for Visual Place Recognition

T0 review · 5 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A decoder-only feature aggregator over a frozen DINOv2 backbone produces global place descriptors that outperform both single-stage and two-stage visual place recognition methods on multiple benchmarks, with roughly one-third the training…

desk verdict Solid empirical VPR paper with a genuinely useful memory-saving adaptation scheme (LoPA); the headline SOTA margins rest on a baseline comparison that needs verification before trusting the exact numbers. read the letter →

arxiv 2412.00784 v2 pith:JDFIZTJQ submitted 2024-12-01 cs.CV

Tong Jin , Feng Lu , Shuyu Hu , Chun Yuan , Yunpeng Liu This is my paper · ORCID
classification cs.CV
keywords visualplacerecognitionfeatureaggregationtransformerdecoderlearnablequerieslow-rankparalleladaptationparameter-efficienttransferlearningDINOv2globaldescriptor
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that a transformer decoder, used purely as a feature aggregator over a frozen foundation-model backbone, can produce place-recognition descriptors that beat both single-stage global-retrieval methods and two-stage methods that add expensive re-ranking. The authors build EDTformer from stacked simplified decoder blocks, self-attention and cross-attention without the feedforward network, where deep features act as keys and values and a set of 64 learnable queries is progressively refined, then compressed through two linear layers into a 4096-dimensional L2-normalized descriptor. A parallel low-rank adapter, LoPA, refines the intermediate features of a frozen DINOv2 backbone so that backpropagation never passes through the backbone, cutting training memory to roughly 5.72 GB. If correct, the practical upshot is that one-stage global retrieval can be both more accurate and considerably cheaper than the current two-stage state of the art.

What carries the argument

The load-bearing mechanism is the simplified decoder block: a self-attention layer followed by a cross-attention layer, with no feedforward network. A fixed set of 64 learnable queries enters the first self-attention layer; in cross-attention the queries attend to the deep features $F$ supplied as keys and values, and after $L=2$ stacked blocks the refined queries are compressed by two linear layers into a 4096-dimensional global descriptor. The supporting mechanism is LoPA (Low-rank Parallel Adaptation): a lightweight parallel network whose adaptation function $h(x)=s\,W_u\,\sigma(W_d x)+x$ takes as input the sum of the previous LoPA output and the current intermediate feature of the frozen DINOv2 block, progressively refining features so that training gradients never backpropagate through the backbone. LoPA contributes only 0.08 million trainable parameters and about 1.74 GB of extra training memory.

What would settle it

Retrain EDTformer and the leading baselines, especially BoQ and SALAD, under one identical protocol, same image size, crop, batch, epochs, and a validation split fixed before any hyperparameter choice, and evaluate on a held-out VPR benchmark not used in ablations; if the reported margins vanish or reverse, the central claim fails. A cheaper falsifier is to measure peak training GPU memory for all four methods at the same batch size, since the paper reports memory at batch 36 while ablations use batch 72.

Watch

Extended reading notes

Core claim

The paper's central claim is that decoder-style aggregation is what has been missing from single-stage VPR. In EDTformer, the image features extracted by the frozen (LoPA-adapted) DINOv2 backbone are linearly projected and then used as keys and values in cross-attention; the queries are learnable parameters, not image-dependent tokens. After two simplified decoder blocks, the learned queries hold the aggregated contextual information and are passed through two fully connected layers, one for dimensionality reduction and one to adjust the number of queries, before flattening and L2 normalization yields the global representation. The authors report that this 4096-dim descriptor achieves state-of-the-art Recall@1 on MSLS-challenge (78.4%), Tokyo24/7 (97.1%), SPED (92.4%), Nordland (88.3%), AmsterTime (65.2%), and the SVOX subsets, exceeding single-stage methods and also the two-stage SelaVPR on Nordland and MSLS, while training in about one-third to half the memory of comparable DINOv2-based aggregators.

Load-bearing premise

The headline comparisons assume that recall numbers reported by different methods are directly comparable, since most baselines are taken from their original papers with different training pipelines, BoQ is re-run with a modified image size, and the architecture choices (two blocks, 64 queries, rank 4) are selected using ablations on the same benchmark sets that later produce the headline results.

Editorial extensions

If this is right

  • A purely global descriptor can outperform two-stage re-ranking pipelines on Nordland and MSLS, so the extra cost of geometric verification or local-feature re-ranking is not required for state-of-the-art accuracy on those benchmarks.
  • Training memory of about 5.72 GB means DINOv2-based VPR can be fine-tuned on a single consumer GPU, widening access to foundation-model-based place recognition.
  • The 4096-dim descriptor can be shrunk to 512 dims by changing only the last linear layer while still beating several published global descriptors, which matters for large-database storage and search speed.
  • Because LoPA improves GeM, NetVLAD, and SALAD as well as EDTformer, the adaptation method is a drop-in efficiency upgrade for existing aggregators.
  • Consistent gains on day/night, seasonal, and domain-shift datasets indicate the decoder queries learn to attend to condition-invariant landmark regions.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • We infer that the same decoder aggregator should transfer to other image-retrieval tasks where a compact global descriptor is desired, since nothing in EDTformer is place-specific beyond the training data.
  • The attention visualizations suggest each query functions as a soft landmark detector; a testable extension is to use the learned query-attention maps as weak labels for unsupervised landmark discovery.
  • If the memory savings really come from avoiding backpropagation through the backbone, LoPA-style parallel adapters could reduce training cost for other foundation-model transfer tasks; this is our inference, since the paper only demonstrates VPR.
  • A cleaner stress test of the reported margins would train BoQ and EDTformer under one identical protocol and would reserve the ablation datasets for a separate validation stage; we infer the margins could shift under such a protocol, though the paper's BoQ rerun already addresses image size.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 5 minor

Summary. The paper proposes EDTformer, a transformer-decoder-based feature aggregator for single-stage visual place recognition (VPR), combined with LoPA, a low-rank parallel adaptation module that keeps a DINOv2 backbone frozen while refining its intermediate features. The global descriptor is formed by a set of learnable queries passed through stacked simplified decoder blocks (self-attention and cross-attention only) followed by two linear layers. The authors report state-of-the-art Recall@N on eight benchmarks, including MSLS, Tokyo24/7, SPED, Nordland, AmsterTime, and SVOX, and claim lower training memory and inference time than comparable DINOv2-based methods. Internal ablations cover the aggregator choice, the adaptation method, the number of decoder blocks and queries, the removal of the FFN, the LoPA rank, and the transferability of LoPA to other aggregators.

Significance. If the empirical claims hold, the paper makes two useful contributions: a decoder-style query-based aggregation that is simpler and more parameter-efficient than BoQ, and a parallel adaptation approach that avoids backpropagation through a frozen foundation model, with a plausible memory advantage over adapter-based PETL. The internal ablations in Tables V-X are broadly consistent and support the main design decisions, and the demonstration that LoPA improves GeM, NetVLAD, and SALAD as well as EDTformer is a genuinely useful result. The central limitation is that the headline SOTA comparison is not a controlled experiment: the only directly re-run baseline (BoQ) is re-run under a modified image-size protocol, several other baselines are quoted from papers with different backbones and training pipelines, and structural hyperparameters are selected on the same evaluation sets later used for the headline tables. These issues affect the claimed margin over prior methods but not the internal consistency of the proposed architecture.

major comments (5)
  1. [Section IV-C, Table II and Table III] The headline claim that EDTformer outperforms BoQ and other SOTA methods is not supported by a controlled comparison. The footnote to Table II states that BoQ is re-run 'strictly following its training pipeline, except for keeping the same image size for both training (224x224) and inference (322x322) as our method.' Since the official BoQ release is not used and no code or checkpoints are provided, the reader cannot verify that this reproduced BoQ is the configuration the community would report; changes in inference resolution can meaningfully alter retrieval accuracy. Other baselines (e.g., SelaVPR with DINOv2-large, SALAD, Patch-NetVLAD, R2Former) are taken from their original papers with different training protocols. The margins cited as evidence (MSLS-challenge 78.4 vs 75.9, Tokyo24/7 97.1 vs 95.2, AmsterTime 65.2 vs 53.6) are therefore not attributable to the proposed method until all methods are evaluated under a common protocol, ideally with released code and checkpoints.
  2. [Section IV-D and Section IV-B] The structural hyperparameters L=2, M=64, and rank=4 are selected by ablation studies on MSLS-val, SPED, Pitts30k, and Nordland (Tables VII, VIII, X), and the same datasets are then used for the headline results in Tables II and III. This selection-on-test-data procedure can inflate the reported margins and makes the comparison to methods whose hyperparameters were not tuned on these benchmark sets unfair. The authors should either use a separated validation set for model selection or explicitly quantify the sensitivity of the final results to the chosen hyperparameters.
  3. [Tables II and III] No error bars, multiple seeds, or significance tests are reported for any of the benchmark numbers. Several of the claimed improvements are small in percentage terms (e.g., 1.5% R@1 over BoQ on Pitts30k, 0.2% over SALAD on MSLS-val), and without repeated runs or a statistical test it is unclear whether these differences are meaningful rather than run-to-run noise. At minimum, the authors should report the mean and standard deviation over at least three training runs for the main comparing methods and for their own method.
  4. [Section I and Section IV-C, two-stage comparison] The claim that EDTformer 'outperforms two-stage VPR methods which add a re-ranking with considerable cost' rests entirely on quoted numbers from Patch-NetVLAD, R2Former, and SelaVPR in Tables II and III. These methods are not re-run under the same training data, backbone, or evaluation pipeline, and the comparison does not report the candidate-set size or re-ranking cost used for the quoted two-stage numbers. Since the stated advantage over two-stage methods is a central selling point of the paper, the authors should provide a controlled head-to-head evaluation (same training data, same query/database splits, same evaluation script) or substantially soften the claim.
  5. [Section IV-B and Section IV-C] The training-memory comparison in Table IV is measured with batch size 36, while the adaptation-method comparison in Table VI is measured with batch size 72. The paper should state why different batch sizes are used and confirm that the relative memory savings are not an artifact of the batch-size choice; ideally, all memory comparisons should be reported at the same batch size or with a scaling analysis.
minor comments (5)
  1. [Section IV-D, Table VI] The text says 'PartialTuning used in BoQ [24] and SALAD [24]' but both citations point to SALAD; the BoQ reference should be [5] and the SALAD reference should be [24].
  2. [Table III] The row labeled 'SelaVPR [24]' should cite [4], not [24]; the same typo appears in the caption context of Table III.
  3. [Fig. 1] The legend contains the typo 'EngenPlaces'; it should read 'EigenPlaces'.
  4. [Tables II and III] The method 'SFRS' is misspelled as 'SRFS' in Tables II and III; the reference list and Section IV use SFRS.
  5. [Eq. (9)] The dimensions in Eq. (9) are not fully explained; in particular, how W3 acts on the transposed matrix to produce the final flattened descriptor should be stated explicitly in the text.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: EDTformer's central claims are empirical benchmark comparisons, not derived quantities that reduce to fitted inputs or self-citations.

full rationale

The paper's core contributions are architectural (EDTformer decoder aggregation, LoPA parallel adaptation) and are evaluated against external benchmarks and external baselines. The aggregation is explicitly defined by equations (6)-(9) from backbone features and learnable queries, and the loss is a standard multi-similarity loss (Eq. 13); no reported Recall@1 value is used as a definitional input to the architecture or as a fitted parameter renamed as a prediction. LoPA is described as 'inspired by previous studies [37]-[39]' (external side-tuning and parallel-tuning works), and the scaling factor is said to be 'As in [3], [4], [38]', but this is a similarity note, not a load-bearing derivation. The self-citations ([3], [4], [57], [69]) are to prior methods used as baselines or related work; they do not supply the benchmark margins or any uniqueness or forcing argument. The BoQ re-run with a modified image size is a baseline-comparability choice, openly disclosed in a footnote, not a fitted input passed off as a prediction. The ablation-based selection of L=2, M=64, and rank=4 on the same evaluation datasets is a benchmark-overfitting risk, not a circular reduction: no equation equates the final reported R@1 to the hyperparameter choice. Therefore the derivation chain is self-contained and no specific circular step can be exhibited.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The central empirical results rest on a moderate set of hand-picked hyperparameters, several selected by ablations on the evaluation benchmarks, and on standard transformer machinery plus assumptions about the transferability of GSV-Cities training and benchmark comparability. No new physical entities are introduced.

free parameters (6)
  • Number of EDTformer decoder blocks (L) = 2
    Selected by ablation on MSLS-val, SPED, Pitts30k and Nordland (Table VII); two blocks give the best overall trade-off and more blocks degrade or add cost.
  • Number of learnable queries (M) = 64
    Selected by ablation on the same benchmarks (Table VIII); 64 queries balance capacity and redundancy.
  • LoPA rank (r) = 4
    Selected by ablation (Table X); rank 4 gives the best overall R@1 across the four datasets.
  • LoPA scaling factor (s) = 0.5
    Set in Section IV-B following adapter-style scaling from prior work; no dedicated ablation is reported for this value.
  • Global descriptor dimensionality = 4096
    Hand-chosen output dimension in Section IV-B; ablations with 512 to 2048 dimensions show declining but competitive performance.
  • MS loss hyperparameters (alpha, beta, lambda, margin) = alpha=1, beta=50, lambda=0, margin=0.1
    Adopted from prior work [1], [3], [5] in Section IV-B; not tuned in this paper.
assumptions (4)
  • domain assumption DINOv2's frozen intermediate features contain enough place-discriminative information for a lightweight parallel adapter to refine them.
    LoPA only processes outputs of each encoder block without modifying the backbone (Section III-C, Eqs. 11-12); if these intermediate features are not informative for place identity, the central performance gain disappears.
  • domain assumption Training on GSV-Cities with multi-similarity loss transfers to the evaluation benchmarks.
    The model is trained on GSV-Cities and evaluated on Pitts30k, MSLS, Tokyo24/7, SPED, Nordland, AmsterTime and SVOX (Section IV-A/B); the comparison assumes this transfer is valid for all methods.
  • domain assumption The evaluation protocol, including thresholds and database/query splits, matches prior published benchmark results.
    Recall@N thresholds are adopted from previous works (Section IV-A) and baseline numbers are quoted from original papers; the BoQ reproduction with modified image size (Section IV-C footnote) shows protocol equivalence cannot be taken for granted.
  • standard math Multi-head scaled dot-product attention is a valid operation for aggregating semantic information into query vectors.
    The EDTformer relies entirely on standard attention equations (3)-(5); no new mathematical machinery is introduced.

how reviews work

0 comments
Cite this review

Pith. "Pith review of EDTformer: An Efficient Decoder Transformer for Visual Place Recognition." pith.science (2026). https://pith.science/paper/JDFIZTJQ

@misc{pith2026241200784,
  author       = {Pith},
  title        = {Pith review of: EDTformer: An Efficient Decoder Transformer for Visual Place Recognition},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JDFIZTJQ}},
  note         = {Machine review of arXiv:2412.00784}
}
read the original abstract

Visual place recognition (VPR) aims to determine the general geographical location of a query image by retrieving visually similar images from a large geo-tagged database. To obtain a global representation for each place image, most approaches typically focus on the aggregation of deep features extracted from a backbone through using current prominent architectures (e.g., CNNs, MLPs, pooling layer, and transformer encoder), giving little attention to the transformer decoder. However, we argue that its strong capability to capture contextual dependencies and generate accurate features holds considerable potential for the VPR task. To this end, we propose an Efficient Decoder Transformer (EDTformer) for feature aggregation, which consists of several stacked simplified decoder blocks followed by two linear layers to directly produce robust and discriminative global representations. Specifically, we do this by formulating deep features as the keys and values, as well as a set of learnable parameters as the queries. Our EDTformer can fully utilize the contextual information within deep features, then gradually decode and aggregate the effective features into the learnable queries to output the global representations. Moreover, to provide more powerful deep features for EDTformer and further facilitate the robustness, we use the foundation model DINOv2 as the backbone and propose a Low-rank Parallel Adaptation (LoPA) method to enhance its performance in VPR, which can refine the intermediate features of the backbone progressively in a memory- and parameter-efficient way. As a result, our method not only outperforms single-stage VPR methods on multiple benchmark datasets, but also outperforms two-stage VPR methods which add a re-ranking with considerable cost. Code will be available at https://github.com/Tong-Jin01/EDTformer.

Figures

Figures reproduced from arXiv: 2412.00784 by the authors.

Figure 1
Figure 1. The performance comparison (Recall@1) on multiple benchmark [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Our pipeline to produce the robust and discriminative global representation for single-stage VPR. Firstly, the frozen backbone with Low-rank Parallel [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Illustration of our Low-rank Parallel Adaptation method. (a) is a standard transformer encoder block in ViT. (b) is the popular PETL method based [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: The R@1 and inference time comparison of different single-stage [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Qualitative results. In these challenging scenarios, our method successfully retrieves the correct images, while other methods commonly return the [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: Visualization of learnable queries. The first column (a) with a green border shows the original input images, the middle 2-5 columns (b) with an orange [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

70 extracted references · 43 canonical work pages

  1. [1]

    MixVPR: Feature mixing for visual place recognition,

    A. Ali-bey, B. Chaib-draa, and P. Gigu `ere, “MixVPR: Feature mixing for visual place recognition,” in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision , 2023, pp. 2998–3007

    work page 2023

  2. [2]

    Eigenplaces: Training viewpoint robust models for visual place recognition,

    G. Berton, G. Trivigno, B. Caputo, and C. Masone, “Eigenplaces: Training viewpoint robust models for visual place recognition,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 11 080–11 090

    2023

  3. [3]

    Cricavpr: Cross-image correlation-aware representation learning for visual place recognition,

    F. Lu, X. Lan, L. Zhang, D. Jiang, Y . Wang, and C. Yuan, “Cricavpr: Cross-image correlation-aware representation learning for visual place recognition,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , June 2024

    work page 2024

  4. [4]

    Towards seamless adaptation of pre-trained models for visual place recognition,

    F. Lu, L. Zhang, X. Lan, S. Dong, Y . Wang, and C. Yuan, “Towards seamless adaptation of pre-trained models for visual place recognition,” in The Twelfth International Conference on Learning Representations , 2024

    work page 2024

  5. [5]

    BoQ: A place is worth a bag of learnable queries,

    A. Ali-bey, B. Chaib-draa, and P. Gigu `ere, “BoQ: A place is worth a bag of learnable queries,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , June 2024, pp. 17 794–17 803

    2024

  6. [6]

    Mapillary street-level sequences: A dataset for lifelong place JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 13 recognition,

    F. Warburg, S. Hauberg, M. Lopez-Antequera, P. Gargallo, Y . Kuang, and J. Civera, “Mapillary street-level sequences: A dataset for lifelong place JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 13 recognition,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2020, pp. 2626–2635

    work page 2021

  7. [7]

    24/7 place recognition by view synthesis,

    A. Torii, R. Arandjelovic, J. Sivic, M. Okutomi, and T. Pajdla, “24/7 place recognition by view synthesis,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2015, pp. 1808– 1817

    work page 2015

  8. [8]

    Vpr-bench: An open-source visual place recognition eval- uation framework with quantifiable viewpoint and appearance change,

    M. Zaffar, S. Garg, M. Milford, J. Kooij, D. Flynn, K. McDonald-Maier, and S. Ehsan, “Vpr-bench: An open-source visual place recognition eval- uation framework with quantifiable viewpoint and appearance change,” International Journal of Computer Vision (IJCV) , pp. 1–39, 2021

    work page 2021

Show all 70 references
  1. [9]

    Amstertime: A visual place recognition benchmark dataset for severe domain shift,

    B. Yildiz, S. Khademi, R. M. Siebes, and J. Van Gemert, “Amstertime: A visual place recognition benchmark dataset for severe domain shift,” in 2022 26th International Conference on Pattern Recognition (ICPR) . IEEE, 2022, pp. 2749–2755

    2022

  2. [10]

    Adaptive- attentive geolocalization from few queries: A hybrid approach,

    G. M. Berton, V . Paolicelli, C. Masone, and B. Caputo, “Adaptive- attentive geolocalization from few queries: A hybrid approach,” in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2021, pp. 2918–2927

    2021

  3. [11]

    Deep visual geo-localization benchmark,

    G. Berton, R. Mereu, G. Trivigno, C. Masone, G. Csurka, T. Sattler, and B. Caputo, “Deep visual geo-localization benchmark,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 5396–5407

    2022

  4. [12]

    Simultaneous localization and mapping: A survey of current trends in autonomous driving,

    G. Bresson, Z. Alsayed, L. Yu, and S. Glaser, “Simultaneous localization and mapping: A survey of current trends in autonomous driving,” IEEE Transactions on Intelligent Vehicles, vol. 2, no. 3, pp. 194–220, 2017

    2017

  5. [13]

    Probabilistic visual place recognition for hierarchical localization,

    M. Xu, N. S ¨underhauf, and M. Milford, “Probabilistic visual place recognition for hierarchical localization,” IEEE Robotics and Automation Letters, vol. 6, no. 2, pp. 311–318, 2020

    2020

  6. [14]

    Global visual and semantic observations for outdoor robot localization,

    J. Li, S. Hu, Q. Li, J. Chen, V . C. M. Leung, and H. Song, “Global visual and semantic observations for outdoor robot localization,” IEEE Transactions on Network Science and Engineering , vol. 8, no. 4, pp. 2909–2921, 2021

    2021

  7. [15]

    A low-cost and scalable framework to build large-scale localization benchmark for augmented reality,

    H. Liu, L. Zhao, Z. Peng, W. Xie, M. Jiang, H. Zha, H. Bao, and G. Zhang, “A low-cost and scalable framework to build large-scale localization benchmark for augmented reality,” IEEE Transactions on Circuits and Systems for Video Technology , vol. 34, no. 4, pp. 2274– 2288, 2024

    2024

  8. [16]

    Visual place recognition: A survey,

    S. Lowry, N. S ¨underhauf, P. Newman, J. J. Leonard, D. Cox, P. Corke, and M. J. Milford, “Visual place recognition: A survey,” IEEE Trans- actions on Robotics , vol. 32, no. 1, pp. 1–19, 2016

    2016

  9. [17]

    Visual place recognition: A tutorial [tutorial],

    S. Schubert, P. Neubert, S. Garg, M. Milford, and T. Fischer, “Visual place recognition: A tutorial [tutorial],” IEEE Robotics & Automation Magazine, vol. 31, no. 3, pp. 139–153, 2024

    2024

  10. [18]

    Unifying deep local and global features for image search,

    B. Cao, A. Araujo, and J. Sim, “Unifying deep local and global features for image search,” in European Conference on Computer Vision. Springer, 2020, pp. 726–743

    2020

  11. [19]

    Netvlad: Cnn architecture for weakly supervised place recognition,

    R. Arandjelovic, P. Gronat, A. Torii, T. Pajdla, and J. Sivic, “Netvlad: Cnn architecture for weakly supervised place recognition,” in IEEE/CVF International Conference on Computer Vision and Pattern Recognition (CVPR), 2016, pp. 5297–5307

    2016

  12. [20]

    Fine-tuning cnn image retrieval with no human annotation,

    F. Radenovi ´c, G. Tolias, and O. Chum, “Fine-tuning cnn image retrieval with no human annotation,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 41, no. 7, pp. 1655–1668, 2018

    2018

  13. [21]

    Patch-netvlad: Multi-scale fusion of locally-global descriptors for place recognition,

    S. Hausler, S. Garg, M. Xu, M. Milford, and T. Fischer, “Patch-netvlad: Multi-scale fusion of locally-global descriptors for place recognition,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 14 141–14 152

    2021

  14. [22]

    Hybrid cnn-transformer features for visual place recognition,

    Y . Wang, Y . Qiu, P. Cheng, and J. Zhang, “Hybrid cnn-transformer features for visual place recognition,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 33, no. 3, pp. 1109–1122, 2023

    2023

  15. [23]

    Rethinking visual geo- localization for large-scale applications,

    G. Berton, C. Masone, and B. Caputo, “Rethinking visual geo- localization for large-scale applications,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022, pp. 4878–4888

    2022

  16. [24]

    Optimal transport aggregation for visual place recognition,

    S. Izquierdo and J. Civera, “Optimal transport aggregation for visual place recognition,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , June 2024

    2024

  17. [25]

    Attention is all you need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” in Advances in neural information processing systems , vol. 30, 2017

    2017

  18. [26]

    Encoder-decoder with cascaded crfs for semantic segmentation,

    J. Ji, R. Shi, S. Li, P. Chen, and Q. Miao, “Encoder-decoder with cascaded crfs for semantic segmentation,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 31, no. 5, pp. 1926–1938, 2021

    1926

  19. [27]

    Efficient self-ensemble framework for semantic segmentation,

    W. Bousselham, G. Thibault, L. Pagano, A. Machireddy, J. Gray, Y . H. Chang, and X. Song, “Efficient self-ensemble framework for semantic segmentation,” arXiv preprint arXiv:2111.13280 , 2021

    2021 arXiv

  20. [28]

    Per-pixel classification is not all you need for semantic segmentation,

    B. Cheng, A. Schwing, and A. Kirillov, “Per-pixel classification is not all you need for semantic segmentation,” Advances in neural information processing systems, vol. 34, pp. 17 864–17 875, 2021

    2021

  21. [29]

    Masked-attention mask transformer for universal image segmentation,

    B. Cheng, I. Misra, A. G. Schwing, A. Kirillov, and R. Girdhar, “Masked-attention mask transformer for universal image segmentation,” arXiv, 2021

    2021

  22. [30]

    Dinov2: Learning robust visual features without supervision,

    M. Oquab, T. Darcet, T. Moutakanni, H. V o, M. Szafraniec, V . Khalidov, P. Fernandez, D. Haziza, F. Massa, A. El-Nouby et al. , “Dinov2: Learning robust visual features without supervision,” arXiv preprint arXiv:2304.07193, 2023

    2023 arXiv

  23. [31]

    Florence: A new foundation model for computer vision,

    L. Yuan, D. Chen, Y .-L. Chen, N. Codella, X. Dai, J. Gao, H. Hu, X. Huang, B. Li, C. Li et al. , “Florence: A new foundation model for computer vision,” arXiv preprint arXiv:2111.11432 , 2021

    2021 arXiv

  24. [32]

    Learning transferable visual models from natural language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark et al., “Learning transferable visual models from natural language supervision,” in International conference on machine learning . PMLR, 2021, pp. 8748–8763

    2021

  25. [33]

    Anyloc: Towards universal visual place recognition,

    N. Keetha, A. Mishra, J. Karhade, K. M. Jatavallabhula, S. Scherer, M. Krishna, and S. Garg, “Anyloc: Towards universal visual place recognition,” arXiv preprint arXiv:2308.00688 , 2023

    2023 arXiv

  26. [34]

    Parameter-efficient transfer learning for nlp,

    N. Houlsby, A. Giurgiu, S. Jastrzebski, B. Morrone, Q. De Laroussilhe, A. Gesmundo, M. Attariyan, and S. Gelly, “Parameter-efficient transfer learning for nlp,” in International Conference on Machine Learning . PMLR, 2019, pp. 2790–2799

    2019

  27. [35]

    Adapt- former: Adapting vision transformers for scalable visual recognition,

    S. Chen, C. Ge, Z. Tong, J. Wang, Y . Song, J. Wang, and P. Luo, “Adapt- former: Adapting vision transformers for scalable visual recognition,” arXiv preprint arXiv:2205.13535 , 2022

    2022 arXiv

  28. [36]

    Convolutional bypasses are better vision transformer adapters,

    S. Jie and Z.-H. Deng, “Convolutional bypasses are better vision transformer adapters,” arXiv preprint arXiv:2207.07039 , 2022

    2022 arXiv

  29. [37]

    Lst: Ladder side-tuning for parameter and memory efficient transfer learning,

    Y .-L. Sung, J. Cho, and M. Bansal, “Lst: Ladder side-tuning for parameter and memory efficient transfer learning,” Advances in Neural Information Processing Systems , vol. 35, pp. 12 991–13 005, 2022

    2022

  30. [38]

    Time-, memory-and parameter-efficient visual adaptation,

    O.-B. Mercea, A. Gritsenko, C. Schmid, and A. Arnab, “Time-, memory-and parameter-efficient visual adaptation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024

    2024

  31. [39]

    Unipt: Universal parallel tuning for transfer learning with efficient parameter and memory,

    H. Diao, B. Wan, Y . Zhang, X. Jia, H. Lu, and L. Chen, “Unipt: Universal parallel tuning for transfer learning with efficient parameter and memory,” arXiv preprint arXiv:2308.14316 , 2023

    2023 arXiv

  32. [40]

    Distinctive image features from scale-invariant keypoints,

    D. G. Lowe, “Distinctive image features from scale-invariant keypoints,” International journal of computer vision , vol. 60, no. 2, pp. 91–110, 2004

    2004

  33. [41]

    Speeded-up robust features (surf),

    H. Bay, A. Ess, T. Tuytelaars, and L. Van Gool, “Speeded-up robust features (surf),” Computer vision and image understanding , vol. 110, no. 3, pp. 346–359, 2008

    2008

  34. [42]

    Fab-map: Probabilistic localization and mapping in the space of appearance,

    M. Cummins and P. Newman, “Fab-map: Probabilistic localization and mapping in the space of appearance,” The International Journal of Robotics Research, vol. 27, no. 6, pp. 647–665, 2008

    2008

  35. [43]

    Fast and incre- mental method for loop-closure detection using bags of visual words,

    A. Angeli, D. Filliat, S. Doncieux, and J.-A. Meyer, “Fast and incre- mental method for loop-closure detection using bags of visual words,” IEEE transactions on robotics , vol. 24, no. 5, pp. 1027–1037, 2008

    2008

  36. [44]

    Large-scale image retrieval with compressed fisher vectors,

    F. Perronnin, Y . Liu, J. S ´anchez, and H. Poirier, “Large-scale image retrieval with compressed fisher vectors,” in 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. IEEE, 2010, pp. 3384–3391

    2010

  37. [45]

    Aggregating local de- scriptors into a compact image representation,

    H. J ´egou, M. Douze, C. Schmid, and P. P ´erez, “Aggregating local de- scriptors into a compact image representation,” in 2010 IEEE computer society conference on computer vision and pattern recognition . IEEE, 2010, pp. 3304–3311

    2010

  38. [46]

    Deep learning features at scale for visual place recognition,

    Z. Chen, A. Jacobson, N. S ¨underhauf, B. Upcroft, L. Liu, C. Shen, I. Reid, and M. Milford, “Deep learning features at scale for visual place recognition,” in 2017 IEEE international conference on robotics and automation (ICRA) . IEEE, 2017, pp. 3223–3230

    2017

  39. [47]

    Learned contextual feature reweighting for image geolocalization,

    H. J. Kim, E. Dunn, and J.-M. Frahm, “Learned contextual feature reweighting for image geolocalization,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , 2017, pp. 2136–2145

    2017

  40. [48]

    End-to-end learning of deep visual representations for image retrieval,

    A. Gordo, J. Almazan, J. Revaud, and D. Larlus, “End-to-end learning of deep visual representations for image retrieval,” International Journal of Computer Vision , vol. 124, no. 2, pp. 237–254, 2017

    2017

  41. [49]

    Self-supervising fine-grained region similarities for large-scale image localization,

    Y . Ge, H. Wang, F. Zhu, R. Zhao, and H. Li, “Self-supervising fine-grained region similarities for large-scale image localization,” in European conference on computer vision . Springer, 2020, pp. 369– 386

    2020

  42. [50]

    Transvpr: Transformer-based place recognition with multi-level attention aggrega- JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 14 tion,

    R. Wang, Y . Shen, W. Zuo, S. Zhou, and N. Zheng, “Transvpr: Transformer-based place recognition with multi-level attention aggrega- JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 14 tion,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Rec...

    2021

  43. [51]

    Sta-vpr: Spatio-temporal alignment for visual place recognition,

    F. Lu, B. Chen, X.-D. Zhou, and D. Song, “Sta-vpr: Spatio-temporal alignment for visual place recognition,” IEEE Robotics and Automation Letters, vol. 6, no. 3, pp. 4297–4304, 2021

    2021

  44. [52]

    Where is your place, visual place recognition?

    S. Garg, T. Fischer, and M. Milford, “Where is your place, visual place recognition?” arXiv preprint arXiv:2103.06443 , 2021

    2021 arXiv

  45. [53]

    Densernet: Weakly supervised visual localization using multi-scale feature aggrega- tion,

    D. Liu, Y . Cui, L. Yan, C. Mousas, B. Yang, and Y . Chen, “Densernet: Weakly supervised visual localization using multi-scale feature aggrega- tion,” in Proceedings of the AAAI conference on artificial intelligence , vol. 35, no. 7, 2021, pp. 6101–6109

    2021

  46. [54]

    Multi- similarity loss with general pair weighting for deep metric learning,

    X. Wang, X. Han, W. Huang, D. Dong, and M. R. Scott, “Multi- similarity loss with general pair weighting for deep metric learning,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 5022–5030

    2019

  47. [55]

    Sinkhorn distances: Lightspeed computation of optimal transport,

    M. Cuturi, “Sinkhorn distances: Lightspeed computation of optimal transport,” in Advances in Neural Information Processing Systems , vol. 26, 2013

    2013

  48. [56]

    R2former: Unified retrieval and reranking transformer for place recognition,

    S. Zhu, L. Yang, C. Chen, M. Shah, X. Shen, and H. Wang, “R2former: Unified retrieval and reranking transformer for place recognition,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 19 370–19 380

    2023

  49. [57]

    Deep homography estimation for visual place recognition,

    F. Lu, S. Dong, L. Zhang, B. Liu, X. Lan, D. Jiang, and C. Yuan, “Deep homography estimation for visual place recognition,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 38, no. 9, 2024, pp. 10 341–10 349

    2024

  50. [58]

    Multi-task learning with multi-query transformer for dense prediction,

    Y . Xu, X. Li, H. Yuan, Y . Yang, and L. Zhang, “Multi-task learning with multi-query transformer for dense prediction,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 34, no. 2, pp. 1228– 1240, 2024

    2024

  51. [59]

    Few-shot learning meets transformer: Unified query-support transformers for few-shot classifica- tion,

    X. Wang, X. Wang, B. Jiang, and B. Luo, “Few-shot learning meets transformer: Unified query-support transformers for few-shot classifica- tion,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 33, no. 12, pp. 7789–7802, 2023

    2023

  52. [60]

    End-to-end object detection with transformers,

    N. Carion, F. Massa, G. Synnaeve, N. Usunier, A. Kirillov, and S. Zagoruyko, “End-to-end object detection with transformers,” in ECCV, 2020, pp. 213–229

    2020

  53. [61]

    Masked-attention mask transformer for universal image segmentation,

    B. Cheng, I. Misra, A. G. Schwing, A. Kirillov, and R. Girdhar, “Masked-attention mask transformer for universal image segmentation,” in 2022 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2022

    2022

  54. [62]

    Image aesthetics assessment via learnable queries,

    Z. Xiong and O. Authors, “Image aesthetics assessment via learnable queries,” in ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2024, pp. 2805–2809

    2024

  55. [63]

    Lora: Low-rank adaptation of large language models,

    E. J. Hu, Y . Shen, P. Wallis, Z. AllenZhu, Y . Li, S. Wang, L. Wang, and W. Chen, “Lora: Low-rank adaptation of large language models,” arXiv preprint arXiv:2106.09685, 2021

    2021 arXiv

  56. [64]

    An image is worth 16x16 words: Transformers for image recognition at scale,

    A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly et al., “An image is worth 16x16 words: Transformers for image recognition at scale,” in International Conference on Learning Representations , 2020

    2020

  57. [65]

    Gaussian error linear units (gelus),

    D. Hendrycks and K. Gimpel, “Gaussian error linear units (gelus),” arXiv preprint arXiv:1606.08415, 2016

    2016 arXiv

  58. [66]

    Gsv-cities: Toward appro- priate supervised visual place recognition,

    A. Ali-bey, B. Chaib-draa, and P. Gigu `ere, “Gsv-cities: Toward appro- priate supervised visual place recognition,” Neurocomputing, vol. 513, pp. 194–203, 2022

    2022

  59. [67]

    Visual place recognition with repetitive structures,

    A. Torii, J. Sivic, T. Pajdla, and M. Okutomi, “Visual place recognition with repetitive structures,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2013, pp. 883–890

    2013

  60. [68]

    Single-view place recognition under seasonal changes,

    D. Olid, J. M. F ´acil, and J. Civera, “Single-view place recognition under seasonal changes,” arXiv preprint arXiv:1808.06516 , 2018

    2018 arXiv

  61. [69]

    Su- pervlad: Compact and robust image descriptors for visual place recogni- tion,

    F. Lu, X. Zhang, C. Ye, S. Dong, L. Zhang, X. Lan, and C. Yuan, “Su- pervlad: Compact and robust image descriptors for visual place recogni- tion,” Advances in Neural Information Processing Systems , vol. 37, pp. 5789–5816, 2024

    2024

  62. [70]

    Only look once, mining distinctive landmarks from convnet for visual place recognition,

    Z. Chen, F. Maffra, I. Sa, and M. Chli, “Only look once, mining distinctive landmarks from convnet for visual place recognition,” in2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2017, pp. 9–16

    2017

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.