Pith. sign in

REVIEW 3 major objections 5 minor 34 references

Environmental Change Detection: Toward a Practical Task of Scene Change Detection

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper claims that scene change detection can be performed from a database of unaligned past images, and that a retrieval-alignment-aggregation framework reaches 0.4815 average F1, surpassing the 0.4291 VPR+SCD baseline and…

desk verdict A genuinely new task formulation for change detection with a solid but slightly over-sold evaluation; the benchmark is curated from known sequences, so 'uncurated' is a stretch. read the letter →

arxiv 2506.11481 v1 pith:XTV5EC5T submitted 2025-06-13 cs.CV

Kyusik Cho , Suhan Woo , Hongje Seong , Euntai Kim This is my paper · ORCID
classification cs.CV
keywords environmentalchangedetectionscenevisualplacerecognitionreferencedatabaseretrievalDINOv2featurespseudo-alignedviewmulti-scalespatialalignmentcross-attentionsemanticaggregation
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 introduces Environmental Change Detection (ECD), a reformulation of scene change detection in which each query image is compared not with one aligned before image but with a large, uncurated database of reference images from nearby viewpoints. It argues that the conventional SCD assumptions of known query-reference pairs and identical viewpoints are unrealistic, and that a system can instead reconstruct the past scene from multiple retrieved references. The proposed framework retrieves top-K references with a visual place recognition model, builds multi-scale pseudo-aligned views via patch matching in DINO feature space, and aggregates them with cross-attention before a change-segmentation head. On three standard SCD benchmarks reconstructed for ECD, the method reaches 0.4815 average F1 at database stride 1, surpassing the VPR+SCD baseline at 0.4291 and approaching the oracle that uses ground-truth aligned references at 0.4997.

What carries the argument

The load-bearing machinery is a two-stage reconstruction: a spatial aligner partitions the query feature into n by n grids, runs a stride-1 sliding-window cosine-similarity search over the DINOv2 feature maps of all retrieved references, and copies the best-matching patch from any reference into a pseudo-aligned view; a semantic aggregator then uses multi-head cross-attention with that pseudo-aligned view as the query and all retrieved reference features as keys and values, averaging across grid resolutions of 1 by 1, 2 by 2, and 4 by 4 to form a reconstructed scene that is fed with the query into a cross-attention change-detection head.

What would settle it

Run the stride-1 evaluation after removing, for each query, every reference image that is not a coarse match, meaning more than 25 meters or 45 degrees away on ChangeSim and a different sequence on VL-CMU-CD or PSCD. If the average F1 gap over the baseline collapses, the result depends on residual near-viewpoint coverage rather than on the aligner's ability to reconstruct arbitrary viewpoints.

Watch

Extended reading notes

Core claim

The central claim is that relabeling scene change detection from compare-a-query-to-its-aligned-reference to answer-a-query-from-a-database-of-unaligned-past-images is tractable, and that the right inductive bias is to separate place-level retrieval, view-level alignment, and scene-level aggregation rather than to force a single spatially matched pair. Concretely, at database stride 1 the framework's average F1 is 0.4815, beating the strong baseline of 0.4291 and coming within 0.018 of the oracle at 0.4997; on ChangeSim and PSCD it even edges out the oracle at stride 1. The authors interpret this as evidence that environmental context from multiple references can compensate for missing exact alignment.

Load-bearing premise

The method assumes that even after striding, the reference database contains enough images with genuine field-of-view overlap to each query for VPR retrieval and DINO patch matching to reconstruct the query viewpoint; without that overlap, there is nothing to align or aggregate.

Editorial extensions

If this is right

  • At database stride 1, ECD reaches 0.4815 average F1, within 0.018 of the oracle, so fully aligned references are not required for near-oracle change detection on these benchmarks.
  • Using more than one reference image helps until irrelevant images are included, so reference diversity matters up to a point rather than raw database size.
  • The ablations show that both the spatial aligner and the semantic aggregator contribute independently, meaning the gain is not simply from having extra reference images.
  • At larger database strides the method stays ahead of the baseline, for example VL-CMU-CD F1 of 0.4355 versus 0.3948 at stride 10, indicating some robustness to sparser reference databases.
  • On ChangeSim, where the ground-truth pairs are only coarsely aligned, the method surpasses the oracle at stride 1, suggesting that retrieved references can sometimes be better aligned than the provided pair.

Reading between the lines

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

  • An untested extension is to replace the frozen DINOv2 feature space with a text-aligned or geometry-aware feature space; the pipeline's retrieval-alignment-aggregation structure does not depend on DINOv2 specifically, so the aligner could inherit stronger viewpoint invariance from the backbone.
  • The VPR analysis in Appendix B suggests that the real bottleneck at high stride is retrieval coverage rather than fusion: strict match accuracy falls to between 13 and 42 percent at stride 10 while coarse match stays near 68 to 98 percent. A testable extension is to couple the framework with a map-based candidate generator or learned reranker to lift coarse-match coverage, which should transfer di
  • On coarsely aligned benchmarks the framework beats the oracle, which implies that retrieval-based reference reconstruction could be used as training-time augmentation for SCD models rather than only as a replacement task.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper introduces Environmental Change Detection (ECD), a variant of Scene Change Detection in which no query-reference pairs are predefined and references come from a database of unaligned images. The proposed framework retrieves the top-K reference images using a frozen BoQ VPR model, constructs multi-scale pseudo-aligned views by patch-level matching in a frozen DINOv2 feature space, aggregates these views with a cross-attention semantic aggregator, and feeds the result into an RSCD-style change detection head. Experiments are conducted on VL-CMU-CD, PSCD, and ChangeSim, reconstructed as ECD benchmarks by striding the original reference database with strides 1, 3, 5, and 10. The method consistently outperforms a BoQ+RSCD baseline across all settings and approaches oracle performance at stride 1, with ablations showing both proposed components contribute.

Significance. If the results hold, ECD is a useful and practical reformulation of change detection, and the proposed retrieval-alignment-aggregation framework is a credible first solution. Strengths of the paper include the clear problem definition, the use of frozen foundation models with a relatively simple trainable module, the cross-dataset evaluation (PSCD is test-only), the retrieval analysis in Appendix B, and the detailed ablations. The main risk is the gap between the abstract's claims of a 'large-scale database of uncurated images' and 'comparable performance to the oracle' on one hand, and the actual benchmark construction and stride-dependent results on the other.

major comments (3)
  1. [Abstract and Section 5.3] The claim of 'achieving comparable performance to the oracle setting' is only supported at database stride 1, where the average F1 is 0.4815 versus the oracle 0.4997. At stride 10 the average F1 drops to 0.3581, a relative gap of about 28% with respect to the same oracle value, and at stride 3 the gap is about 17.5%. The abstract and Section 5.3 should either qualify the oracle-comparability claim by stride or restrict it to the most favorable setting.
  2. [Sections 3.2, 6, and Appendix A.1/B] The abstract describes the reference database as a 'large-scale database of uncurated images', but the benchmark is constructed by striding the original SCD sequences, which preserves strong field-of-view overlap. Appendix B shows coarse-match retrieval rates of 98.10% (ChangeSim), 83.74% (VL-CMU-CD), and 68.29% (PSCD) even at stride 10, and Section 6 concedes that 'the reference database must include FOV overlaps with queries to enable change detection'. The experiments therefore demonstrate a curated-but-unaligned retrieval task, not a genuinely uncurated database. Please soften the claim or add an experiment with distractor/unrelated images to test robustness to truly uncurated references.
  3. [Tables 1-4] No error bars or multiple seeds are reported, and several differences are small (e.g., PSCD oracle 0.3370 versus ours 0.3540 at stride 1; K=3 versus K=5 in Table 3 at stride 1: 0.4815 versus 0.4871). The claim that the method 'consistently outperforms' the baseline would be substantially stronger if results were reported as the mean and standard deviation over at least three seeds.
minor comments (5)
  1. [Appendix A.2] The sentence 'its architecture is identical to the PSCD [11]' should refer to RSCD [11], not PSCD.
  2. [Equation (3)] The expression 'arg max_{k,q}' uses q both as the query image and as a spatial location; it should be 'arg max_{k,z}' or equivalent notation.
  3. [Figure 7 caption] The caption states that 'the pseudo-aligned view at fine grid levels works robustly', but the preceding sentence says fine-grid alignment may fail under large object changes; the intended wording is presumably 'coarse grid levels'.
  4. [References] References [13] and [16] are duplicates of the same work, as are [14] and [17]; these should be consolidated.
  5. [Section 5.1] The PSCD description says 'only evaluations are performed with the best model trained and validated on VL-CMU-CD'; the wording is unclear and should specify that PSCD is used exclusively as a test set without fine-tuning.

Circularity Check

0 steps flagged · score 2.0 of 10

No load-bearing circularity: the ECD pipeline is tested against held-out labels and an oracle bound, with the sole self-citation appearing only as related work.

full rationale

The paper's central claim—that retrieval, multi-scale patch alignment, and semantic aggregation can detect scene changes from an unaligned reference database—is evaluated on held-out test splits of VL-CMU-CD, PSCD, and ChangeSim. The method parameters (K=3, hierarchical grid sizes) are selected from analysis experiments, not fitted to test labels, and the oracle row is an upper bound computed with ground-truth references rather than a prediction derived from the model. The only author self-citation is [18] in Related Work, where it is mentioned as a prior zero-shot SCD method; no equation, benchmark construction, or central argument depends on it. Appendix A.1's database striding is a benchmark-design choice, and Appendix B's coarse-match statistics quantify the residual overlap rather than assume it. Section 6 explicitly concedes that the reference database must include FOV overlaps with queries; this limits the claimed generality to truly uncurated databases but does not make the derivation circular. None of the paper's equations reduce to fitted parameters or to inputs by definition, and no prediction is constructed from the quantity it claims to predict. The observed gap between the abstract's 'uncurated' framing and the curated-but-strided benchmark is a scope limitation, not circularity.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claim rests on standard deep-learning transfer assumptions: frozen pretrained features (DINOv2, BoQ) generalize to the evaluation datasets; the reconstructed labels from the original SCD datasets remain valid when the reference view is unaligned; and the strided database preserves enough spatial overlap. The two explicit hyperparameters selected by experiment are K=3 and the multi-scale grid configuration.

free parameters (2)
  • Number of reference candidates K = 3
    Chosen by analysis experiments (Table 3); default K=3 balances accuracy vs computation.
  • Multi-scale grid resolutions for spatial aligner = 1x1, 2x2, 4x4 averaged (Hier. x3)
    Selected via ablation (Table 4); hierarchical x3 gives best average F1.
assumptions (4)
  • domain assumption VPR retrieval (BoQ) returns at least one reference image with spatial overlap for each query
    Section 4.1 constructs R from TopKVPR; Appendix B quantifies retrieval accuracy, showing strict matches can be rare at high strides.
  • domain assumption DINOv2 features support cross-view patch correspondence at the feature level
    Section 4.2 uses frozen DINO features for sliding-window patch matching; no alignment supervision is used.
  • domain assumption Original SCD ground-truth change masks remain valid when the reference is a different, unaligned image
    Section 5.1 reuses the SCD labels y as-is for ECD, even though the reference view differs from the query.
  • domain assumption Benchmark striding creates a realistic ECD setting
    Section 5.2 applies stride s per sequence to simulate missing aligned references; realism is not validated against external data.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Environmental Change Detection: Toward a Practical Task of Scene Change Detection." pith.science (2026). https://pith.science/paper/XTV5EC5T

@misc{pith2026250611481,
  author       = {Pith},
  title        = {Pith review of: Environmental Change Detection: Toward a Practical Task of Scene Change Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XTV5EC5T}},
  note         = {Machine review of arXiv:2506.11481}
}
read the original abstract

Humans do not memorize everything. Thus, humans recognize scene changes by exploring the past images. However, available past (i.e., reference) images typically represent nearby viewpoints of the present (i.e., query) scene, rather than the identical view. Despite this practical limitation, conventional Scene Change Detection (SCD) has been formalized under an idealized setting in which reference images with matching viewpoints are available for every query. In this paper, we push this problem toward a practical task and introduce Environmental Change Detection (ECD). A key aspect of ECD is to avoid unrealistically aligned query-reference pairs and rely solely on environmental cues. Inspired by real-world practices, we provide these cues through a large-scale database of uncurated images. To address this new task, we propose a novel framework that jointly understands spatial environments and detects changes. The main idea is that matching at the same spatial locations between a query and a reference may lead to a suboptimal solution due to viewpoint misalignment and limited field-of-view (FOV) coverage. We deal with this limitation by leveraging multiple reference candidates and aggregating semantically rich representations for change detection. We evaluate our framework on three standard benchmark sets reconstructed for ECD, and significantly outperform a naive combination of state-of-the-art methods while achieving comparable performance to the oracle setting. The code will be released upon acceptance.

Figures

Figures reproduced from arXiv: 2506.11481 by the authors.

Figure 1
Figure 1. Comparison between the problem settings of SCD and ECD. In the conventional SCD setting, each query image is provided with its corresponding reference image as a predefined pair. In contrast, ECD considers a more challenging scenario where query-reference pair are unknown. ECD just provides a database of unaligned reference images. Furthermore, ECD does not guarantee the presence of a perfectly matched pair within t… view at source ↗
Figure 2
Figure 2. Overview. We first construct a subset of the reference database containing images relevant to the query using the VPR module. Then, the spatial aligner focuses on the spatial information of the features to generate a pseudo-aligned view that matches the perspective of the query image. Based on this pseudo-aligned view, the semantic aggregator integrates the semantic information across the subset. Finally, both the r… view at source ↗
Figure 3
Figure 3. Illustration of spatial aligner and semantic aggregator. (a) The input query feature fq is divided into an n × n grid. For each grid cell, we perform a sliding window search with stride 1 over all reference features {f (k) r } K k=1 to find the best-matching patch. By collecting the best matches across all grid cells, we construct a pseudo-aligned reference representation ˜fr,n that approximates the viewpoint of the… view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Qualitative results. The baseline struggles with viewpoint variation, often predicting changes at incorrect locations. In contrast, our method is more robust to viewpoint mismatch. Note that r (oracle) is not included in the reference database but is shown for illustra…
Figure 5
Figure 5. Figure 5: Example images from the reference database. The figure compares the full reference database used in the conventional SCD setting with the strided reference database in the proposed ECD setup. The images outlined in red boxes represent the reference images selected for …
Figure 6
Figure 6. Figure 6: Examples of pseudo-aligned views. Given a query image q, the spatial aligner generates pseudo-aligned views at multiple grid levels using reference images r (1) , r (2), and r (3) . r (oracle) is shown for reference purposes. At finer grid levels, the generated views c…
Figure 7
Figure 7. Figure 7: Examples of pseudo-aligned views under large object changes. When a substantial change spans the entire area of a patch, the pseudo-aligned view at fine grid levels may fail to align accurately. However, the pseudo-aligned view at fine grid levels works robustly, provi…
Figure 8
Figure 8. Figure 8: Examples of pseudo-aligned views under a large database stride. Even when the reference subset R contains many incorrect reference images, the spatial aligner effectively suppresses the influence of unrelated images and reconstructs the pseudo-aligned view primarily us…
Figure 9
Figure 9. Figure 9: Qualitative results on ChangeSim. Our method performs robust predictions even in complex scenes by effectively integrating spatial information. 20 [PITH_FULL_IMAGE:figures/full_fig_p020_9.png]
Figure 10
Figure 10. Figure 10: Qualitative results on VL-CMU-CD. Our method is more robust to viewpoint variations than the baseline. 21 [PITH_FULL_IMAGE:figures/full_fig_p021_10.png]
Figure 11
Figure 11. Figure 11: Qualitative results on PSCD. Even when the VPR retrieval includes incorrect matches, our model can still produce accurate predictions. 22 [PITH_FULL_IMAGE:figures/full_fig_p022_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 29 canonical work pages

  1. [1]

    Street- view change detection with deconvolutional networks

    Pablo F Alcantarilla, Simon Stent, German Ros, Roberto Arroyo, and Riccardo Gherardi. Street- view change detection with deconvolutional networks. Autonomous Robots, 42:1301–1322, 2018

    work page 2018

  2. [2]

    Weakly supervised silhouette-based semantic scene change detection

    Ken Sakurada, Mikiya Shibuya, and Weimin Wang. Weakly supervised silhouette-based semantic scene change detection. In 2020 IEEE International conference on robotics and automation (ICRA), pages 6861–6867. IEEE, 2020

    work page 2020

  3. [3]

    Emplace: Self-supervised urban scene change detection

    Tim Alpherts, Sennay Ghebreab, and Nanne van Noord. Emplace: Self-supervised urban scene change detection. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 1737–1745, 2025

    work page 2025

  4. [4]

    Change detection from a street image pair using cnn features and superpixel segmentation

    Ken Sakurada and Takayuki Okatani. Change detection from a street image pair using cnn features and superpixel segmentation. In British Machine Vision Conference (BMVC), 2015

    work page 2015

  5. [5]

    Generalizable disaster damage assessment via change detection with vision foundation model

    Kyeongjin Ahn, Sungwon Han, Sungwon Park, Jihee Kim, Sangyoon Park, and Meeyoung Cha. Generalizable disaster damage assessment via change detection with vision foundation model. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 27784–27792, 2025

    work page 2025

  6. [6]

    Umad: University of macau anomaly detection benchmark dataset

    Dong Li, Lineng Chen, Cheng-Zhong Xu, and Hui Kong. Umad: University of macau anomaly detection benchmark dataset. In 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 5836–5843. IEEE, 2024

    work page 2024

  7. [7]

    Changesim: Towards end-to-end online scene change detection in industrial indoor environments

    Jin-Man Park, Jae-Hyuk Jang, Sahng-Min Yoo, Sun-Kyung Lee, Ue-Hwan Kim, and Jong- Hwan Kim. Changesim: Towards end-to-end online scene change detection in industrial indoor environments. In 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 8578–8585. IEEE, 2021

    work page 2021

  8. [8]

    Dual task learning by leveraging both dense correspondence and mis-correspondence for robust change detection with imperfect matches

    Jin-Man Park, Ue-Hwan Kim, Seon-Hoon Lee, and Jong-Hwan Kim. Dual task learning by leveraging both dense correspondence and mis-correspondence for robust change detection with imperfect matches. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 13749–13759, 2022

    work page 2022

Show all 34 references
  1. [9]

    Hierarchical paired channel fusion network for street scene change detection

    Yinjie Lei, Duo Peng, Pingping Zhang, Qiuhong Ke, and Haifeng Li. Hierarchical paired channel fusion network for street scene change detection. IEEE Transactions on Image Processing, 30:55–67, 2020

    2020

  2. [10]

    How to reduce change detection to semantic segmentation

    Guo-Hua Wang, Bin-Bin Gao, and Chengjie Wang. How to reduce change detection to semantic segmentation. Pattern Recognition, 138:109384, 2023

    2023

  3. [11]

    Robust scene change de- tection using visual foundation models and cross-attention mechanisms

    Chun-Jung Lin, Sourav Garg, Tat-Jun Chin, and Feras Dayoub. Robust scene change de- tection using visual foundation models and cross-attention mechanisms. arXiv preprint arXiv:2409.16850, 2024

    2024 arXiv

  4. [12]

    Self- supervised pretraining for scene change detection

    Vijaya Raghavan T Ramkumar, Prashant Bhat, Elahe Arani, and Bahram Zonooz. Self- supervised pretraining for scene change detection. In Proceedings of the 35th Conference on Neural Information Processing Systems, pages 6–14, 2021

    2021

  5. [13]

    The change you want to see

    Ragav Sachdeva and Andrew Zisserman. The change you want to see. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pages 3993–4002, 2023

    2023

  6. [14]

    Semi-supervised scene change detection by distilla- tion from feature-metric alignment

    Seonhoon Lee and Jong-Hwan Kim. Semi-supervised scene change detection by distilla- tion from feature-metric alignment. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), pages 1226–1235, 2024. 10

    2024

  7. [15]

    Dr-tanet: Dynamic receptive temporal atten- tion network for street scene change detection

    Shuo Chen, Kailun Yang, and Rainer Stiefelhagen. Dr-tanet: Dynamic receptive temporal atten- tion network for street scene change detection. In 2021 IEEE Intelligent Vehicles Symposium (IV), pages 502–509. IEEE, 2021

    2021

  8. [16]

    The change you want to see

    Ragav Sachdeva and Andrew Zisserman. The change you want to see. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2023

    2023

  9. [17]

    Semi-supervised scene change detection by distilla- tion from feature-metric alignment

    Seonhoon Lee and Jong-Hwan Kim. Semi-supervised scene change detection by distilla- tion from feature-metric alignment. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), pages 1226–1235, January 2024

    2024

  10. [18]

    Zero-shot scene change detection

    Kyusik Cho, Dong Yeop Kim, and Euntai Kim. Zero-shot scene change detection. In Pro- ceedings of the AAAI Conference on Artificial Intelligence , volume 39, pages 2509–2517, 2025

    2025

  11. [19]

    Netvlad: Cnn architecture for weakly supervised place recognition

    Relja Arandjelovic, Petr Gronat, Akihiko Torii, Tomas Pajdla, and Josef Sivic. Netvlad: Cnn architecture for weakly supervised place recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 5297–5307, 2016

    2016

  12. [20]

    Object recognition from local scale-invariant features

    David G Lowe. Object recognition from local scale-invariant features. In Proceedings of the seventh IEEE international conference on computer vision, volume 2, pages 1150–1157. Ieee, 1999

    1999

  13. [21]

    Emvp: Embracing visual foundation model for visual place recognition with centroid-free probing

    Qibo Qiu, Shun Zhang, Haiming Gao, Honghui Yang, Haochao Ying, Wenxiao Wang, and Xiaofei He. Emvp: Embracing visual foundation model for visual place recognition with centroid-free probing. Advances in Neural Information Processing Systems, 37:120928–120950, 2024

    2024

  14. [22]

    Supervlad: Compact and robust image descriptors for visual place recognition

    Feng Lu, Xinyao Zhang, Canming Ye, Shuting Dong, Lijun Zhang, Xiangyuan Lan, and Chun Yuan. Supervlad: Compact and robust image descriptors for visual place recognition. Advances in Neural Information Processing Systems, 37:5789–5816, 2024

    2024

  15. [23]

    R2former: Unified retrieval and reranking transformer for place recognition

    Sijie Zhu, Linjie Yang, Chen Chen, Mubarak Shah, Xiaohui Shen, and Heng Wang. R2former: Unified retrieval and reranking transformer for place recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 19370– 19380, 2023

    2023

  16. [24]

    Transvpr: Transformer-based place recognition with multi-level attention aggregation

    Ruotong Wang, Yanqing Shen, Weiliang Zuo, Sanping Zhou, and Nanning Zheng. Transvpr: Transformer-based place recognition with multi-level attention aggregation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pages 13648–13657, 2022

    2022

  17. [25]

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

    Feng Lu, Xiangyuan Lan, Lijun Zhang, Dongmei Jiang, Yaowei Wang, and Chun Yuan. Cricavpr: Cross-image correlation-aware representation learning for visual place recognition. In Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 16...

    2024

  18. [26]

    Optimal transport aggregation for visual place recognition

    Sergio Izquierdo and Javier Civera. Optimal transport aggregation for visual place recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 17658–17668, 2024

    2024

  19. [27]

    Dinov2: Learning robust visual features without supervision

    Maxime Oquab, Timothée Darcet, Théo Moutakanni, Huy V o, Marc Szafraniec, Vasil Khalidov, Pierre Fernandez, Daniel Haziza, Francisco Massa, Alaaeldin El-Nouby, et al. Dinov2: Learning robust visual features without supervision. arXiv preprint arXiv:2304.07193, 2023

    2023 arXiv

  20. [28]

    Boq: A place is worth a bag of learnable queries

    Amar Ali-Bey, Brahim Chaib-draa, and Philippe Giguère. Boq: A place is worth a bag of learnable queries. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 17794–17803, 2024

    2024

  21. [29]

    Zeroscd: Zero-shot street scene change detection

    Shyam Sundar Kannan and Byung-Cheol Min. Zeroscd: Zero-shot street scene change detection. arXiv preprint arXiv:2409.15255, 2024. 11

    2024 arXiv

  22. [30]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. Advances in neural information processing systems, 30, 2017

    2017

  23. [31]

    Kingma and Jimmy Ba

    Diederik P. Kingma and Jimmy Ba. Adam: A method for stochastic optimization. In Interna- tional Conference for Learning Representations, 2015

    2015

  24. [32]

    Sgdr: Stochastic gradient descent with warm restarts

    Ilya Loshchilov and Frank Hutter. Sgdr: Stochastic gradient descent with warm restarts. In International Conference for Learning Representations, 2017

    2017

  25. [33]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016

    2016

  26. [34]

    Bench- marking image retrieval for visual localization

    Noé Pion, Martin Humenberger, Gabriela Csurka, Yohann Cabon, and Torsten Sattler. Bench- marking image retrieval for visual localization. In 2020 International Conference on 3D Vision (3DV), pages 483–494. IEEE, 2020. 12 A Implementation details A.1 Details on ECD setup prepar...

    2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.