Ranking-aware Continual Learning for LiDAR Place Recognition

Gengxuan Tian; Junqiao Zhao; Qiankun Yu; Qiwen Gu; Siyue Tao; Tiantian Feng; Xufei Wang

REVIEW 2 major objections 5 minor 1 cited by

Ranking-aware Continual Learning for LiDAR Place Recognition

T0 review · 2 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read KDF claims LiDAR place-recognition models can learn new environments without forgetting old ones if knowledge distillation preserves retrieval ranking and the old model's features are fused with the new model's at test time.

desk verdict Useful distillation idea for continual LiDAR place recognition, but the fusion module makes the main forgetting comparison unfair; needs a baseline that isolates distillation from old-model concatenation. read the letter →

arxiv 2505.07198 v1 pith:KTACNTWT submitted 2025-05-12 cs.CV

Xufei Wang , Gengxuan Tian , Junqiao Zhao , Siyue Tao , Qiwen Gu , Qiankun Yu , Tiantian Feng This is my paper · ORCID

classification cs.CV

keywords continuallearningLiDARplacerecognitionknowledgedistillationcatastrophicforgettingfusionranking-awarelosspointcloudretrievalSLAM

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

The paper is trying to establish that a retrieval model can learn new places without forgetting old ones if, during training, it is forced to preserve not just feature similarity but the order in which old and new models rank candidate places, and if, at test time, the old model's knowledge is explicitly kept alongside the new model's. The authors build a continual-learning wrapper, KDF, around any LiDAR place-recognition encoder. They show it reduces catastrophic forgetting on a four-dataset sequential benchmark and improves generalization to an unseen city, across four different encoder backbones. If true, place-recognition systems in SLAM and autonomous driving could be updated incrementally as vehicles enter new environments, instead of being retrained from scratch.

What carries the argument

The central machinery is the soft ranking function $R(q,i)=1+\sum_{j\neq i} G((S(q,j)-S(q,i))/\tau)$, a Smooth-AP-style differentiable proxy for the rank of candidate $i$ against query $q$. KDF computes this ranking for embeddings from the old and new encoders and minimizes their absolute difference, then adds a symmetric KL divergence between the two embedding distributions; place-recognition metric learning with a triplet margin loss runs on top. A separate knowledge fusion module then concatenates old and new descriptors into a fused descriptor used at retrieval time.

What would settle it

Run the four-step protocol but evaluate using only the new model's descriptor at test time, without concatenating the old model's features; if mean Recall@1 and forgetting score then fall to the no-fusion ablation level or below InCloud and CCL, the claim that the learned model itself resists forgetting is unsupported.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that catastrophic forgetting in LiDAR place recognition is not a fixed cost of sequential training: it can be cut sharply by aligning the old and new models on the quantity the retrieval task actually cares about, the ranking of candidate matches, and by refusing to discard the old model. The paper reports KDF improving mean Recall@1 over InCloud by 5.56 points and lowering the forgetting score by 9.93 points when averaged across four encoder backbones under the triplet-loss protocol, with the best single backbone (MinkLoc3D) reaching a forgetting score of 1.20. On an unseen KITTI evaluation, the paper reports KDF achieving the best overall Recall@1 on sequences 02, 07, and 08 among the compared methods.

Load-bearing premise

The method's forgetting gains assume that the previous model stays frozen and available at test time so its features can be concatenated with the new model's; if a system cannot store that old model, most of the reported improvement disappears, and the paper does not spell out how the fused representation is maintained after a third environment is added.

Editorial extensions

If this is right

A robot or vehicle entering a new city can update its place-recognition model on new LiDAR data while keeping high Recall@1 on previously visited cities, without storing the full old training set.
The ranking-aware distillation loss can be dropped into existing descriptors such as PointNetVLAD, LoGG3D-Net, MinkLoc3D, and TransLoc3D, so continual place recognition does not require a new architecture.
Because retrieval is the actual downstream task, preserving the ranking of candidate matches transfers the knowledge that matters for place recognition rather than only matching intermediate feature statistics.
The fused old-plus-new descriptor improves generalization: after sequential training, the model scores higher on an unseen dataset than several baselines, suggesting the old model contributes transferable knowledge.
The framework's components are separable: ablations show ranking distillation, symmetric-KL distribution distillation, and fusion each add to mean Recall@1 and reduce forgetting on MinkLoc3D.

Reading between the lines

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

Beyond the paper, a deployment-time descriptor that concatenates every old model grows linearly with the number of environments; a practical system would need a pruning or re-distillation policy that KDF does not specify.
Beyond the paper, the ranking-distillation idea should transfer to visual and multimodal place recognition, since the retrieval objective is identical; the paper lists visual place recognition as future work.
Beyond the paper, the current experiments fuse only the two most recent models, so how KDF behaves on a third sequential environment is untested; training on a fifth domain and checking early-domain Recall@1 would settle it.

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.

Desk Editor's Note

read the letter

First, the bottom line: the paper's main claim is undercut by an unfair evaluation setup. The knowledge fusion module concatenates the frozen old model's descriptor with the new model's at test time (Eq. 11). Since the old model never changes, it retains its old-domain knowledge perfectly, so the forgetting score (Eq. 12) on the fused system is largely predetermined. A trivial baseline that fine-tunes without any distillation and concatenates old and new features would likely produce low forgetting too. The paper never reports such a control, so you cannot attribute the improvements to the ranking-aware distillation.

That said, the ranking-aware distillation idea is genuinely new and plausible. Adapting Smooth AP to preserve retrieval ranking across domains is a reasonable way to transfer place-recognition knowledge. The paper also does a broad empirical job: four backbones, four datasets, several baselines including recent MICL, and a KITTI generalization test. The ablation shows the distillation losses help even without fusion: forgetting drops from 7.99 to 5.17 on MinkLoc3D. So there is a real effect, just not the one advertised.

The soft spots are concentrated in the evaluation. Besides the fusion confound, the multi-step protocol is underspecified. After training on a third environment, do you keep both prior descriptors? If you keep only the immediately previous model, early-task knowledge is lost; if you keep all, the dimension grows and the training procedure is not described. The distribution distillation loss (Eq. 7) applies KL divergence to embeddings without explaining how vectors become distributions. Also, no code, no error bars, and key hyperparameters (tau, m) are missing.

The citation pattern is fine; it engages with the relevant prior work.

Overall, this paper is for a niche audience of continual-LPR researchers. The ranking-aware loss is worth exploring, but the current manuscript needs a major revision. The authors should compare against a like-for-like baseline that either uses only the new model at test time or allows all methods the same old-model concatenation. They also need to spell out the protocol for more than two environments.

I'd still send it to peer review, because there is a real idea here and the experiments are broad enough to deserve referee time. But I'd expect the reviewers to demand the control experiment and a rewritten evaluation section.

Reading group: maybe, as a case study in evaluation pitfalls.

Referee Report

2 major / 5 minor

Summary. The paper proposes KDF, a continual learning framework for LiDAR place recognition. It introduces a ranking-aware knowledge distillation loss that combines a ranking-based distillation term with a symmetric KL-divergence distribution-based term, and a knowledge fusion module that concatenates the old and new models' descriptors at test time. Experiments are conducted with a 4-step protocol on Oxford, DCC, Riverside, and In-house using four backbones (PointNetVLAD, LoGG3D-Net, MinkLoc3D, TransLoc3D), reporting mean Recall@1 and forgetting scores, plus KITTI generalization results. The authors claim that KDF can be applied to different networks and surpasses InCloud, CCL, and MICL in mean Recall@1 and forgetting score.

Significance. Ranking-aware distillation is a sensible and potentially useful idea: adapting Smooth-AP-style ranking relaxation to continual LPR is novel, and the loss is cleanly defined and architecture-agnostic. The experimental breadth is a strength: four backbones, two training losses, comparisons with three recent continual LPR methods, and a generalization protocol on KITTI. The ablation study is well structured and confirms that each proposed loss term contributes positively. However, the central forgetting claim is weakened by the evaluation design: the forgetting score is computed on a fused system that includes the frozen old model, so the reported low forgetting is partly an artifact of retaining the old model rather than evidence that the new model retains old-domain knowledge. The paper is potentially publishable after the evaluation is deconfounded and the multi-step fusion protocol is specified.

major comments (2)

[Section IV-C, Eq. (11), Section V-C, Eq. (12)] The forgetting score F is computed on the fused descriptor f_fuse = {f^t ⊕ f^{t+1}}, which includes the frozen old model's descriptor. Since the old model has already seen the old domains, the fused system's Recall@1 on old tasks is inflated by the old model's retained knowledge; the new model's own forgetting is not measured. The paper's own ablation (Table V) shows that adding Knowledge Fusion to the full distillation configuration reduces F from 5.17 to 1.20, and fusion alone reduces F from 7.99 to 4.64, so the fusion module, rather than the ranking-aware distillation, drives most of the reported forgetting improvement. Please report Recall@1 and F for the new model alone (without fusion), and add a control baseline that fine-tunes without any KD loss but concatenates the frozen-old and new descriptors at test time; without this control, the contribution of ranking-aware distillation to the forgetting reduction is unverified.
[Section IV-C and Section V-A] The paper does not specify how the fused representation is maintained across the four sequential environments. At step 3, if the "old model" is the model after step 2, the step-1 model is discarded; if instead the fused descriptor from step 2 is used as the old knowledge, the training procedure for that fused feature extractor is not described. Consequently, the reported 4-step forgetting scores are not reproducible, and the claim that early-task knowledge is retained beyond two steps is unverified. Please define the exact fusion protocol for steps t=2,3,4, including which model is frozen and what is concatenated at test time for each step.

minor comments (5)

[Section IV-A, Eq. (5)] The denominator N_b^3 in Eq. (5) appears inconsistent with the summation over N_b^2 pairs (q,i); as written, the ranking distillation loss magnitude scales inversely with batch size, which is particularly relevant because the paper uses batch expansion. If the normalization is intentional, please clarify the scaling rationale.
[Section IV, Eqs. (5)-(11)] The notation switches between E^t, E^{t+1}, f^t, and f^{t+1} without clarifying the relationship; please define the embedding notation once and use it consistently throughout.
[Table V] The ablation table uses checkmarks without explicit row configuration labels; as presented, the reader cannot unambiguously map each row to the exact set of enabled components. Please enumerate the configurations.
[Section V-E, last paragraph] The claim that "KDF achieves the best performance on the 02, 07, and 08 sequences" is not accurate for the triplet variant on sequence 08 (Ours: 26.10 vs MICL-T: 29.33 and MICL-C: 32.71); it is Ours-C that is best on 08. Please report per-variant results precisely and avoid conflating the two variants.
[Section V-D, Table II] Several comparisons in Table II are within a fraction of a percentage point (e.g., Ours-C vs CCL in mean Recall@1), and no variance or multiple-seed results are reported. Please report standard deviations or at least clarify how many runs the reported numbers are based on.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: KDF's gains are empirical; the fusion confound is a comparison-fairness gap, not a by-construction identity.

full rationale

The paper's claimed derivation chain — ranking-aware knowledge distillation, symmetric-KL distribution distillation, and triplet metric learning jointly reduce catastrophic forgetting — is not circular. The ranking loss (Eqs. 2-5) minimizes a smooth surrogate of pairwise ranking differences between old and new embeddings, which is a regularizer defined independently of the final evaluation metric. The distribution loss (Eqs. 6-7) is a standard symmetric KL divergence. Neither loss is fitted to the test set or to the forgetting score. Knowledge fusion (Eq. 11) concatenates the frozen old model's descriptor with the new model's descriptor, and the forgetting score (Eq. 12) is computed on that fused system; this makes the old model's retained knowledge able to inflate R_{T,t}. That is a genuine evaluation confound and a fair-comparison risk against single-encoder baselines, and the paper does not include a freeze-and-concatenate control without distillation. It also does not specify how the fusion is maintained across more than two sequential environments. However, this is not circularity: Eq. 11 does not define forgetting, the old model used at the final step is not the original task-1 model but a sequentially updated model, and the ablation in Table V shows a fusion-free variant (Base+LRKD+LDKD, F=5.17) that already improves on the InCloud baseline (F=5.99) on MinkLoc3D. The distillation component therefore has independent empirical support. No self-citation chain or imported uniqueness theorem is load-bearing: the cited prior continual-LPR methods [13], [14], [38] are external baselines. The central claim rests on experiments rather than on a definitional identity, so no step reduces to its own input by construction.

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

The central claim rests on standard continual learning premises plus the ad hoc assumption that storing the old model at test time is acceptable. No invented entities are introduced. The free parameters are hyperparameters not reported.

free parameters (2)

soft-rank temperature tau = not reported
Used in the sigmoid G(x;tau) in Eq. 4 for ranking-based distillation; value not specified in implementation details.
triplet margin m = not reported
Margin in the metric learning loss of Eq. 8; value not reported.

assumptions (4)

domain assumption Knowledge distillation from a frozen old model preserves old-domain discriminative structure
Core premise of the distillation losses; supported by prior continual learning literature.
ad hoc to paper Retaining the old model at test time is allowed and does not violate continual learning constraints
The knowledge fusion module concatenates old and new features (Eq. 11), requiring the old model to be stored and used at inference.
domain assumption Soft ranking computed from pairwise Euclidean similarities is a meaningful distillation target for place recognition
Based on Smooth AP [21] and retrieval practice.
domain assumption The 4-step protocol datasets (Oxford, DCC, Riverside, In-house) are representative of distinct environments
Evaluation protocol inherited from InCloud/CCL.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Ranking-aware Continual Learning for LiDAR Place Recognition." pith.science (2026). https://pith.science/paper/KTACNTWT

@misc{pith2026250507198,
  author       = {Pith},
  title        = {Pith review of: Ranking-aware Continual Learning for LiDAR Place Recognition},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KTACNTWT}},
  note         = {Machine review of arXiv:2505.07198}
}

read the original abstract

Place recognition plays a significant role in SLAM, robot navigation, and autonomous driving applications. Benefiting from deep learning, the performance of LiDAR place recognition (LPR) has been greatly improved. However, many existing learning-based LPR methods suffer from catastrophic forgetting, which severely harms the performance of LPR on previously trained places after training on a new environment. In this paper, we introduce a continual learning framework for LPR via Knowledge Distillation and Fusion (KDF) to alleviate forgetting. Inspired by the ranking process of place recognition retrieval, we present a ranking-aware knowledge distillation loss that encourages the network to preserve the high-level place recognition knowledge. We also introduce a knowledge fusion module to integrate the knowledge of old and new models for LiDAR place recognition. Our extensive experiments demonstrate that KDF can be applied to different networks to overcome catastrophic forgetting, surpassing the state-of-the-art methods in terms of mean Recall@1 and forgetting score.

Figures

Figures reproduced from arXiv: 2505.07198 by the authors.

**Figure 2.** An overview of our proposed method KDF. We first initialize a new model from the old model and freeze the old model. Next, memory exemplars [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** The ranking-based knowledge distillation loss [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: Detailed performance comparison in the 4-step protocol across PointNetVLAD, LoGG3D-Net, MinkLoc3D, and TransLoc3D with methods InCloud [13], CCL [14], and MICL [38]. The results are composed of Recall@1 matrices, where each row represents a continual learning method an…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Place Recognition Meet Multiple Modalitie: A Comprehensive Review, Current Challenges and Future Directions
cs.CV 2025-05 reject novelty 4.0 of 10

A survey of visual, LiDAR, and cross-modal place recognition with a unified code library, but riddled with errors and disclaimer-ridden experimental comparisons.

Reference graph

Works this paper leans on

44 extracted references · 28 canonical work pages · cited by 1 Pith paper

[1]

A survey on global lidar localization: Challenges, advances and open problems,

H. Yin, X. Xu, S. Lu, X. Chen, R. Xiong, S. Shen, C. Stachniss, and Y . Wang, “A survey on global lidar localization: Challenges, advances and open problems,” International Journal of Computer Vision , pp. 1– 33, 2024

work page 2024
[2]

General Place Recognition Survey: Towards Real-World Autonomy

P. Yin, J. Jiao, S. Zhao, L. Xu, G. Huang, H. Choset, S. Scherer, and J. Han, “General place recognition survey: Towards real-world autonomy,” arXiv preprint arXiv:2405.04812 , 2024

work page Pith review arXiv 2024
[3]

Lidar-based place recognition for au- tonomous driving: A survey,

P. Shi, Y . Zhang, and J. Li, “Lidar-based place recognition for au- tonomous driving: A survey,” arXiv preprint arXiv:2306.10561 , 2023

arXiv 2023
[4]

Pointnetvlad: Deep point cloud based retrieval for large-scale place recognition,

M. A. Uy and G. H. Lee, “Pointnetvlad: Deep point cloud based retrieval for large-scale place recognition,” inProceedings of the IEEE conference on computer vision and pattern recognition , 2018, pp. 4470–4479

work page 2018
[5]

Minkloc3d: Point cloud based large-scale place recog- nition,

J. Komorowski, “Minkloc3d: Point cloud based large-scale place recog- nition,” in Proceedings of the IEEE/CVF Winter Conference on Appli- cations of Computer Vision , 2021, pp. 1790–1799

work page 2021
[6]

Logg3d-net: Locally guided global descriptor learning for 3d place recognition,

K. Vidanapathirana, M. Ramezani, P. Moghadam, S. Sridharan, and C. Fookes, “Logg3d-net: Locally guided global descriptor learning for 3d place recognition,” in 2022 International Conference on Robotics and Automation (ICRA). IEEE, 2022, pp. 2215–2221

work page 2022
[7]

Improving point cloud based place recognition with ranking-based loss and large batch training,

J. Komorowski, “Improving point cloud based place recognition with ranking-based loss and large batch training,” in 2022 26th International Conference on Pattern Recognition (ICPR) , 2022, pp. 3699–3705

work page 2022
[8]

Lpd-net: 3d point cloud learning for large-scale place recognition and environment analysis,

Z. Liu, S. Zhou, C. Suo, P. Yin, W. Chen, H. Wang, H. Li, and Y .-H. Liu, “Lpd-net: 3d point cloud learning for large-scale place recognition and environment analysis,” in Proceedings of the IEEE/CVF international conference on computer vision , 2019, pp. 2831–2840

work page 2019

Show all 44 references

[9]

Overlapnet: A siamese network for computing lidar scan similarity with applications to loop closing and localization,

X. Chen, T. L ¨abe, A. Milioto, T. R ¨ohling, J. Behley, and C. Stachniss, “Overlapnet: A siamese network for computing lidar scan similarity with applications to loop closing and localization,” Autonomous Robots, pp. 1–21, 2022

2022
[10]

Transloc3d: point cloud based large-scale place recognition using adaptive receptive fields,

T.-X. Xu, Y .-C. Guo, Z. Li, G. Yu, Y .-K. Lai, and S.-H. Zhang, “Transloc3d: point cloud based large-scale place recognition using adaptive receptive fields,” Communications in Information and Systems , vol. 23, no. 1, pp. 57–83, 2023

2023
[11]

Bioslam: A bioinspired lifelong memory system for general place recognition,

P. Yin, A. Abuduweili, S. Zhao, L. Xu, C. Liu, and S. Scherer, “Bioslam: A bioinspired lifelong memory system for general place recognition,” IEEE Transactions on Robotics , 2023

2023
[12]

A comprehensive survey of continual learning: Theory, method and application,

L. Wang, X. Zhang, H. Su, and J. Zhu, “A comprehensive survey of continual learning: Theory, method and application,” IEEE Transactions on Pattern Analysis and Machine Intelligence , 2024

2024
[13]

Incloud: Incremental learning for point cloud place recognition,

J. Knights, P. Moghadam, M. Ramezani, S. Sridharan, and C. Fookes, “Incloud: Incremental learning for point cloud place recognition,” in 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2022, pp. 8559–8566

2022
[14]

Ccl: Continual contrastive learning for lidar place recognition,

J. Cui and X. Chen, “Ccl: Continual contrastive learning for lidar place recognition,” IEEE Robotics and Automation Letters , vol. 8, no. 8, pp. 4433–4440, 2023

2023
[15]

A fast histogram-based similarity measure for detecting loop closures in 3-d lidar data,

T. R ¨ohling, J. Mack, and D. Schulz, “A fast histogram-based similarity measure for detecting loop closures in 3-d lidar data,” in 2015 IEEE/RSJ international conference on intelligent robots and systems (IROS) . IEEE, 2015, pp. 736–741

2015
[16]

Scan context: Egocentric spatial descriptor for place recognition within 3d point cloud map,

G. Kim and A. Kim, “Scan context: Egocentric spatial descriptor for place recognition within 3d point cloud map,” in 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2018, pp. 4802–4809

2018
[17]

Intensity scan context: Coding intensity and geometry relations for loop closure detection,

H. Wang, C. Wang, and L. Xie, “Intensity scan context: Coding intensity and geometry relations for loop closure detection,” in 2020 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2020, pp. 2095–2101

2020
[18]

Pointnet: Deep learning on point sets for 3d classification and segmentation,

C. R. Qi, H. Su, K. Mo, and L. J. Guibas, “Pointnet: Deep learning on point sets for 3d classification and segmentation,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2017, pp. 652–660

2017
[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 Proceed- ings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 5297–5307

2016
[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
[21]

Smooth-ap: Smoothing the path towards large-scale image retrieval,

A. Brown, W. Xie, V . Kalogeiton, and A. Zisserman, “Smooth-ap: Smoothing the path towards large-scale image retrieval,” in European Conference on Computer Vision . Springer, 2020, pp. 677–694

2020
[22]

Continual learning for real-world autonomous systems: Algorithms, challenges and frameworks,

K. Shaheen, M. A. Hanif, O. Hasan, and M. Shafique, “Continual learning for real-world autonomous systems: Algorithms, challenges and frameworks,” Journal of Intelligent & Robotic Systems , vol. 105, no. 1, p. 9, 2022

2022
[23]

Continual learning for robotics: Definition, framework, learning strategies, opportunities and challenges,

T. Lesort, V . Lomonaco, A. Stoian, D. Maltoni, D. Filliat, and N. D ´ıaz- Rodr´ıguez, “Continual learning for robotics: Definition, framework, learning strategies, opportunities and challenges,” Information fusion , vol. 58, pp. 52–68, 2020

2020
[24]

icarl: Incremental classifier and representation learning,

S.-A. Rebuffi, A. Kolesnikov, G. Sperl, and C. H. Lampert, “icarl: Incremental classifier and representation learning,” in Proceedings of the IEEE conference on Computer Vision and Pattern Recognition, 2017, pp. 2001–2010

2017
[25]

Continual learning with deep generative replay,

H. Shin, J. K. Lee, J. Kim, and J. Kim, “Continual learning with deep generative replay,” Advances in neural information processing systems , vol. 30, 2017

2017
[26]

Overcoming catastrophic forgetting in neural networks,

J. Kirkpatrick, R. Pascanu, N. Rabinowitz, J. Veness, G. Desjardins, A. A. Rusu, K. Milan, J. Quan, T. Ramalho, A. Grabska-Barwinska, et al. , “Overcoming catastrophic forgetting in neural networks,” Pro- ceedings of the national academy of sciences , vol. 114, no. 13, pp. 352...

2017
[27]

Continual learning through synaptic intelligence,

F. Zenke, B. Poole, and S. Ganguli, “Continual learning through synaptic intelligence,” in International conference on machine learning. PMLR, 2017, pp. 3987–3995

2017
[28]

Knowledge distillation: A survey,

J. Gou, B. Yu, S. J. Maybank, and D. Tao, “Knowledge distillation: A survey,” International Journal of Computer Vision , vol. 129, no. 6, pp. 1789–1819, 2021

2021
[29]

Relational knowledge distilla- tion,

W. Park, D. Kim, Y . Lu, and M. Cho, “Relational knowledge distilla- tion,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2019, pp. 3967–3976

2019
[30]

Overcoming catas- trophic forgetting with hard attention to the task,

J. Serra, D. Suris, M. Miron, and A. Karatzoglou, “Overcoming catas- trophic forgetting with hard attention to the task,” in International conference on machine learning . PMLR, 2018, pp. 4548–4557

2018
[31]

Packnet: Adding multiple tasks to a single network by iterative pruning,

A. Mallya and S. Lazebnik, “Packnet: Adding multiple tasks to a single network by iterative pruning,” in Proceedings of the IEEE conference on Computer Vision and Pattern Recognition , 2018, pp. 7765–7773

2018
[32]

Learning without forgetting,

Z. Li and D. Hoiem, “Learning without forgetting,” IEEE transactions on pattern analysis and machine intelligence , vol. 40, no. 12, pp. 2935– 2947, 2017

2017
[33]

Incremental few- shot semantic segmentation via embedding adaptive-update and hyper- class representation,

G. Shi, Y . Wu, J. Liu, S. Wan, W. Wang, and T. Lu, “Incremental few- shot semantic segmentation via embedding adaptive-update and hyper- class representation,” in Proceedings of the 30th ACM international conference on multimedia , 2022, pp. 5547–5556

2022
[34]

Unsupervised model adaptation for continual semantic segmentation,

S. Stan and M. Rostami, “Unsupervised model adaptation for continual semantic segmentation,” in Proceedings of the AAAI conference on artificial intelligence, vol. 35, no. 3, 2021, pp. 2593–2601

2021
[35]

Static-dynamic co-teaching for class- incremental 3d object detection,

N. Zhao and G. H. Lee, “Static-dynamic co-teaching for class- incremental 3d object detection,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 36, no. 3, 2022, pp. 3436–3445

2022
[36]

Wanderlust: Online continual object detection in the real world,

J. Wang, X. Wang, Y . Shang-Guan, and A. Gupta, “Wanderlust: Online continual object detection in the real world,” in Proceedings of the IEEE/CVF international conference on computer vision , 2021, pp. 10 829–10 838

2021
[37]

Airloop: Lifelong loop closure de- tection,

D. Gao, C. Wang, and S. Scherer, “Airloop: Lifelong loop closure de- tection,” in 2022 International Conference on Robotics and Automation (ICRA). IEEE, 2022, pp. 10 664–10 671

2022
[38]

Micl: Mutual information guided continual learning for lidar place recognition,

B. Liu, T. Yang, Y . Fang, and Z. Yan, “Micl: Mutual information guided continual learning for lidar place recognition,” IEEE Robotics and Automation Letters , 2024

2024
[39]

On information and sufficiency,

S. Kullback and R. A. Leibler, “On information and sufficiency,” The annals of mathematical statistics , vol. 22, no. 1, pp. 79–86, 1951

1951
[40]

Lifelong person re- identification via knowledge refreshing and consolidation,

C. Yu, Y . Shi, Z. Liu, S. Gao, and J. Wang, “Lifelong person re- identification via knowledge refreshing and consolidation,” in Proceed- ings of the AAAI Conference on Artificial Intelligence , vol. 37, no. 3, 2023, pp. 3295–3303

2023
[41]

1 year, 1000 km: The oxford robotcar dataset,

W. Maddern, G. Pascoe, C. Linegar, and P. Newman, “1 year, 1000 km: The oxford robotcar dataset,” The International Journal of Robotics Research, vol. 36, no. 1, pp. 3–15, 2017

2017
[42]

Mulran: Multimodal range dataset for urban place recognition,

G. Kim, Y . S. Park, Y . Cho, J. Jeong, and A. Kim, “Mulran: Multimodal range dataset for urban place recognition,” in 2020 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2020, pp. 6246–6253

2020
[43]

Vision meets robotics: The kitti dataset,

A. Geiger, P. Lenz, C. Stiller, and R. Urtasun, “Vision meets robotics: The kitti dataset,” The International Journal of Robotics Research , vol. 32, no. 11, pp. 1231–1237, 2013

2013
[44]

Jensen-shannon divergence and hilbert space embedding,

B. Fuglede and F. Topsoe, “Jensen-shannon divergence and hilbert space embedding,” in International symposium onInformation theory, 2004. ISIT 2004. Proceedings. IEEE, 2004, p. 31

2004

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [1]

A survey on global lidar localization: Challenges, advances and open problems,

H. Yin, X. Xu, S. Lu, X. Chen, R. Xiong, S. Shen, C. Stachniss, and Y . Wang, “A survey on global lidar localization: Challenges, advances and open problems,” International Journal of Computer Vision , pp. 1– 33, 2024

work page 2024

[2] [2]

General Place Recognition Survey: Towards Real-World Autonomy

P. Yin, J. Jiao, S. Zhao, L. Xu, G. Huang, H. Choset, S. Scherer, and J. Han, “General place recognition survey: Towards real-world autonomy,” arXiv preprint arXiv:2405.04812 , 2024

work page Pith review arXiv 2024

[3] [3]

Lidar-based place recognition for au- tonomous driving: A survey,

P. Shi, Y . Zhang, and J. Li, “Lidar-based place recognition for au- tonomous driving: A survey,” arXiv preprint arXiv:2306.10561 , 2023

arXiv 2023

[4] [4]

Pointnetvlad: Deep point cloud based retrieval for large-scale place recognition,

M. A. Uy and G. H. Lee, “Pointnetvlad: Deep point cloud based retrieval for large-scale place recognition,” inProceedings of the IEEE conference on computer vision and pattern recognition , 2018, pp. 4470–4479

work page 2018

[5] [5]

Minkloc3d: Point cloud based large-scale place recog- nition,

J. Komorowski, “Minkloc3d: Point cloud based large-scale place recog- nition,” in Proceedings of the IEEE/CVF Winter Conference on Appli- cations of Computer Vision , 2021, pp. 1790–1799

work page 2021

[6] [6]

Logg3d-net: Locally guided global descriptor learning for 3d place recognition,

K. Vidanapathirana, M. Ramezani, P. Moghadam, S. Sridharan, and C. Fookes, “Logg3d-net: Locally guided global descriptor learning for 3d place recognition,” in 2022 International Conference on Robotics and Automation (ICRA). IEEE, 2022, pp. 2215–2221

work page 2022

[7] [7]

Improving point cloud based place recognition with ranking-based loss and large batch training,

J. Komorowski, “Improving point cloud based place recognition with ranking-based loss and large batch training,” in 2022 26th International Conference on Pattern Recognition (ICPR) , 2022, pp. 3699–3705

work page 2022

[8] [8]

Lpd-net: 3d point cloud learning for large-scale place recognition and environment analysis,

Z. Liu, S. Zhou, C. Suo, P. Yin, W. Chen, H. Wang, H. Li, and Y .-H. Liu, “Lpd-net: 3d point cloud learning for large-scale place recognition and environment analysis,” in Proceedings of the IEEE/CVF international conference on computer vision , 2019, pp. 2831–2840

work page 2019