Pith. sign in

REVIEW 2 major objections 5 minor 68 references

Parameter-Efficient Vision-Language Adaptation with Continuous Metadata Conditioning for Animal Re-Identification

T0 review · 2 major / 5 minor · reviewed 2026-07-13 · grok-4.5

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

Pith's one-line read Continuous numerical metadata, used only while training, makes animal re-identification more robust to years of growth and seasonal change without needing that metadata at test time.

desk verdict Solid, reproducible PEFT recipe for longitudinal animal ReID; continuous training-time metadata in the prompt is the real (if modest) novelty and the evidence holds. read the letter →

arxiv 2607.09443 v1 pith:NT35S5OB submitted 2026-07-10 cs.CV cs.AI

Anil Osman Tur , Tonje Knutsen Sordalen , Kim Tallaksen Halvorsen , Cigdem Beyan This is my paper · ORCID
classification cs.CVcs.AI
keywords AnimalRe-IdentificationVision-LanguageModelsPromptLearningLow-RankAdaptationContinuousMetadataConditioningLongitudinalEvaluationCLIP
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

Long-term animal re-identification must match the same individual across months or years even as size, coloration, and life stage drift. This paper argues that a frozen CLIP vision-language backbone can be adapted for that task with far fewer trainable parameters by combining low-rank visual updates, learnable prompt tokens, and cross-modal alignment. Its central claim is that continuous numerical metadata (capture time, body length, year) should be injected directly into the prompt as smooth embeddings during training rather than being discretized into text bins or fused through extra modules. Because the metadata never appears at inference, the deployed system remains a pure visual nearest-neighbor matcher. On a seven-year corkwing-wrasse dataset and several other wildlife benchmarks, the continuous-conditioning strategy improves closed-set, open-set, and time-aware accuracy over prior CLIP-based ReID methods while keeping the model lightweight.

What carries the argument

Continuous metadata-conditioned prompt learning: real-valued attributes are encoded by sinusoidal projection or FiLM modulation and appended as a token to identity-conditioned CLIP prompts during training, then completely removed at inference so that only the LoRA-adapted visual encoder is used for retrieval.

What would settle it

On a held-out multi-year recapture set, retrain the identical architecture but replace continuous metadata injection with random noise or with no metadata at all; if closed-set and time-aware open-set mAP remain statistically indistinguishable from continuous conditioning, the claim that continuous structure is what drives the improvement fails.

Watch

Extended reading notes

Core claim

Preserving the continuous structure of numerical metadata and injecting it into the prompt representation only during training produces a smoother geometry in the shared embedding space that remains useful for pure visual retrieval after every text and metadata component is discarded, improving robustness to longitudinal appearance change and temporal distribution shift.

Load-bearing premise

The assumption that continuous metadata injected only into the frozen text prompt during training will permanently reshape the visual embedding space so that the same gains remain after metadata and all text components are thrown away at test time.

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

2 major / 5 minor

Summary. The paper proposes a parameter-efficient CLIP adaptation framework for longitudinal animal re-identification. It jointly trains LoRA modules in a frozen ViT vision encoder with learnable prompt context tokens and symmetric cross-modal alignment, while introducing continuous metadata conditioning (sinusoidal encoding or FiLM modulation of numerical attributes such as capture year, day-sequence, and body length) that is injected into the prompt representation only during training. All text and metadata components are discarded at inference, yielding a purely visual nearest-neighbor pipeline. The primary claimed contribution is that preserving the continuous structure of numerical metadata, rather than discretizing it into textual bins, induces smooth geometric modulation of the embedding space that improves robustness under identity and temporal distribution shifts. Evaluation uses a seven-year Melops corkwing-wrasse dataset under closed-set, open-set, time-aware closed-set, and time-aware open-set protocols, plus transfer to multiple wildlife benchmarks including SeaTurtleID2022.

Significance. If the empirical claims hold, the work supplies a practical recipe for ecological ReID: metadata available only at training time can still improve a deployable visual model without inference-time dependencies or large architectural additions. The continuous-versus-discrete comparison, multi-protocol design (including explicit temporal separation), dual-space triplet and auxiliary supervision ablations, prompt-layout ensembling, and multi-seed reporting constitute a thorough experimental package. Public code and evaluation splits further raise the contribution's value. Parameter counts (approximately 73 M trainable versus 150 M+ for full CLIP fine-tuning or prior CLIP-ReID/IndivAID pipelines) while matching or exceeding full fine-tuning on Melops closed-set body crops are a concrete efficiency gain. The longitudinal Melops setting itself, with multi-year growth and seasonal change, is a useful stress test for the community.

major comments (2)
  1. Tables 4–5 and Table 14 (TACS/TAOS): the time-aware protocols rest on very small query sets (37 identities / 76 query images for TACS; 86 identities / 173 query images for TAOS) and short observed temporal spans (roughly 1–113 days). While the paper correctly notes that this sparsity is inherent to capture–mark–recapture ecology, the absolute mAP/Rank gains attributed to continuous metadata conditioning under temporal shift therefore rest on limited statistical power. A bootstrap or query-level uncertainty analysis (beyond five random seeds) would make the longitudinal robustness claim more secure.
  2. Sec. 4.5.2 and Tables 6–7: the best-performing metadata attribute and encoding (Year+FiLM versus DaySeq+Sin, single versus multi-attribute) varies by protocol and crop type, and the main comparison tables report a single chosen configuration “for clarity.” The manuscript should state explicitly whether these configurations were selected a priori, by validation, or post hoc, and should report the full range of continuous-conditioning outcomes (or a pre-specified default such as capture year) so that the claimed advantage of continuous over discrete conditioning cannot be read as selective reporting.
minor comments (5)
  1. Sec. 3.7: the sentence ending “…to obtain the final descriptor. the Euclidean distance…” has a capitalization error and should be cleaned.
  2. Figure 3 caption and Sec. 4.5.9: the logistic-regression slopes are reported as ranges; adding the exact fitted coefficients (or a small table) would aid reproducibility of the temporal-decay analysis.
  3. Table 12: “Ours + Meta (All 3 +FiLM)” jumps to 98.20 M parameters; a one-sentence explanation of which FiLM components drive the increase would help readers judge the efficiency claim under the heaviest metadata setting.
  4. Sec. 2.4 and related-work discussion of MetaWild [29]: the discrete textual baseline is a useful control, but a short clarification that the discrete bins were defined on the whole dataset (as stated later) versus training-set only would remove any residual leakage concern.
  5. Throughout: “Re-ID” / “ReID” and “vision–language” / “vision-language” hyphenation is inconsistent; a single house style would improve polish.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: empirical PEFT + training-time conditioning paper whose claims rest on held-out ranking metrics, not on self-definitional or fitted-by-construction predictions.

full rationale

The paper is a standard methodological/empirical CV contribution. Its central claims (continuous metadata conditioning of CLIP prompts during training yields stronger purely visual embeddings under longitudinal/open-set protocols; LoRA + prompt + dual-space losses enable parameter-efficient adaptation) are evaluated by training on one split and measuring mAP/Rank-k on disjoint query/gallery images under CS/OS/TACS/TAOS protocols (Tables 2–7, 13). Metadata (year, day-sequence, length) is injected only into the frozen text prompt via sinusoidal PE or FiLM (Sec. 3.4) and is discarded at inference; nothing is fitted to a target quantity that is then re-reported as a prediction. Ablations (Tables 8–11) and temporal-decay logistic regressions (Fig. 3) are diagnostic, not circular. Self-citations are limited to the authors’ own Melops dataset release ([52,53]), which is independently described and used as one of several benchmarks; the method is also evaluated on external WildlifeDatasets and SeaTurtleID2022. No uniqueness theorems, ansatz smuggling, or renaming of known results appear. The derivation chain therefore does not reduce to its inputs by construction.

Assumptions & free parameters 5 free parameters · 4 assumptions · 2 invented entities

The central claim rests on standard PEFT and CLIP machinery plus the modeling choice that continuous numerical attributes can be linearly or FiLM-modulated into the frozen text embedding space and still shape visual geometry after those components are removed. Hyper-parameters (rank, context length, loss weights) are free; no new physical entities are postulated.

free parameters (5)
  • LoRA rank r and scale α = r=16, α=16
    Chosen as r=16, α=16 after ablation; larger rank (512) collapses performance. Directly controls capacity of visual adaptation.
  • Number of learnable context tokens M = 4
    Set to M=4; affects prompt capacity and is not derived.
  • Loss weights λ_tri, λ_cm and temperature τ = 1
    Fixed to 1 following prior CLIP-ReID practice; not re-tuned but still free design choices that affect the composite objective.
  • Sinusoidal frequency count and FiLM basis tokens M_a = 16 freqs, M_a=n_a
    16 frequencies per attribute and M_a = n_a are hand-chosen; they determine the continuous metadata embedding dimension.
  • Learning rate and weight decay = 5e-4 / 1e-5
    AdamW 5e-4 / 1e-5; standard but free.
assumptions (4)
  • domain assumption Frozen CLIP visual and text encoders already contain transferable identity-discriminative features that can be specialized by low-rank updates and prompt tokens.
    Invoked throughout Sec. 3; standard PEFT premise for CLIP-based ReID.
  • domain assumption Numerical metadata (length, day-sequence, year) are smoothly correlated with gradual morphological and seasonal appearance change.
    Stated in Sec. 1 and 3.4 as biological motivation for continuous rather than discrete conditioning.
  • ad hoc to paper Identity tokens can be fixed random embeddings in text space that condition prompts without becoming learnable class proxies.
    Sec. 3.3 explicitly distinguishes them from proxy-based metric learning; the fixed-token design is a paper-specific modeling choice.
  • domain assumption Symmetric image-to-text and text-to-image contrastive losses plus dual-space triplet losses produce complementary geometry that survives removal of the text branch.
    Sec. 3.6 and ablation 4.5.7; common in CLIP-ReID literature but still an unproved modeling assumption for the longitudinal setting.
invented entities (2)
  • Continuous metadata-conditioned prompt token (sinusoidal or FiLM)
    purpose: Inject real-valued attributes into the frozen CLIP text embedding sequence during training only, so that the visual encoder internalizes longitudinal variation.
    Primary methodological contribution (Sec. 3.4). Independent evidence is the continuous-vs-discrete ablation and the time-aware gains; no external physical prediction is made.
  • Fixed non-learnable identity token in prompt space
    purpose: Provide identity-specific conditioning for cross-modal alignment without acting as a visual prototype.
    Sec. 3.3; ablation shows modest but consistent drop when removed. Not a new particle or force, but a paper-specific representational device.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Parameter-Efficient Vision-Language Adaptation with Continuous Metadata Conditioning for Animal Re-Identification." pith.science (2026). https://pith.science/paper/NT35S5OB

@misc{pith2026260709443,
  author       = {Pith},
  title        = {Pith review of: Parameter-Efficient Vision-Language Adaptation with Continuous Metadata Conditioning for Animal Re-Identification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NT35S5OB}},
  note         = {Machine review of arXiv:2607.09443}
}
read the original abstract

Long-term animal re-identification (ReID) must remain robust to gradual morphological evolution and seasonal appearance shifts. Although recent vision-language models provide strong pretrained visual representations, adapting them to longitudinal ecological settings remains challenging, particularly under identity and temporal distribution shifts. We present a parameter-efficient CLIP adaptation framework for animal ReID and introduce a continuous metadata-conditioning mechanism that incorporates numerical attributes directly into the prompt representation during training. While low-rank visual adaptation, prompt-based supervision, and cross-modal alignment provide the adaptation framework, the proposed metadata-conditioning strategy constitutes the primary methodological contribution. By preserving the continuous structure of numerical metadata rather than discretizing it into textual categories, the proposed approach enables smooth modulation of the embedding space during training while maintaining a purely visual inference pipeline. Experiments on a seven-year longitudinal fish dataset and multiple wildlife benchmarks demonstrate improved performance under closed-set, open-set, and time-aware evaluation protocols. The results demonstrate that continuous metadata conditioning improves robustness to longitudinal appearance variation and temporal distribution shifts, while parameter-efficient adaptation enables a purely visual inference pipeline without requiring metadata at test time. Code and evaluation splits can be found at: https://github.com/AnilOsmanTur/MetaPrompt-ReID.

Figures

Figures reproduced from arXiv: 2607.09443 by the authors.

Figure 1
Figure 1. Overview of the proposed framework. Training (left): A frozen CLIP ViT vision encoder is adapted using LoRA modules inserted into the query, key, and value linear projection layers of each multi-head self-attention block. The visual embedding 𝐟𝑣 and its CLIP-projected counterpart 𝐟 proj 𝑣 are supervised with dual-space batch-hard triplet losses and auxiliary identity classification heads. In parallel, identity-condi… view at source ↗
Figure 2
Figure 2. Multiple observations of the same individual from the Melops dataset [53] across different capture years. The top and middle rows show full-body lateral views from different time points and body sides, illustrating natural growth, size variation, and side-specific appearance differences. The bottom row presents corresponding head crops, highlighting fine-grained facial patterns that remain discriminative despite lon… view at source ↗
Figure 3
Figure 3. Predicted identification probability as a function of temporal distance under the proposed method. Curves represent fitted binomial logistic regression models for Rank-1 and Rank-5 retrieval across evaluation protocols (CS, OS, TACS, and TAOS). Shaded regions denote 95% confidence intervals. Temporal distance is log-transformed as log(1 + days). Prediction ranges are restricted to the empirically observed temporal s… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: ReID performance across model configurations, stratified by temporal context and body side. Points represent the proportion of queries for which the correct individual was retrieved within the top candidate identities (Rank-1 and Rank-5), with whiskers indicating 95% b…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

68 extracted references · 4 canonical work pages

  1. [1]

    Sea- turtleid2022: A long-span dataset for reliable sea turtle re- identification, in: IEEE WACV

    Adam, L., Čermák, V., Papafitsoros, K., Picek, L., 2024. Sea- turtleid2022: A long-span dataset for reliable sea turtle re- identification, in: IEEE WACV

    2024

  2. [2]

    Transformer-based models for enhanced amur tiger re-identification, in: Proc

    Bai, X., Islam, T., Bin Azhar, M., 2024. Transformer-based models for enhanced amur tiger re-identification, in: Proc. of IEEE SAMI

    2024

  3. [3]

    From species-specific models to universal re-id: a survey of animal re-identification

    Beyan, C., Osman Tur, A., Karimi, E., 2026. From species-specific models to universal re-id: a survey of animal re-identification. In- formation Fusion 133, 104323. URL:https://www.sciencedirect. com/science/article/pii/S1566253526002022,doi:https://doi.org/10. 1016/j.inffus.2026.104323

    arXiv 2026

  4. [4]

    Metric learning for prototype-based classification, in: Innovations in Neural Information Paradigms and Applications

    Biehl, M., Hammer, B., Schneider, P., Villmann, T., 2009. Metric learning for prototype-based classification, in: Innovations in Neural Information Paradigms and Applications. Springer, pp. 183–199

    2009

  5. [5]

    A purely visual re-id approach for bumblebees and its application to ecological mon- itoring

    Borlinghaus, P., Tausch, F., Rettenberger, L., 2023. A purely visual re-id approach for bumblebees and its application to ecological mon- itoring. Smart Agricultural Technology

    2023

  6. [6]

    Trajectory-basedfisheventclassification through pre-training with diffusion models

    Canovi, N., Ellis, B.A., Sørdalen, T.K., Allken, V., Halvorsen, K.T., Malde,K.,Beyan,C.,2024. Trajectory-basedfisheventclassification through pre-training with diffusion models. Ecological Informatics 82, 102733

    2024

  7. [7]

    Wildfusion:Individualanimalidentificationwithcalibratedsimilarity fusion

    Cermak, V., Picek, L., Adam, L., Neumann, L., Matas, J., 2024. Wildfusion:Individualanimalidentificationwithcalibratedsimilarity fusion. arXiv preprint arXiv:2408.12934

    arXiv 2024

  8. [8]

    Detection features as attention (defat): A keypoint-free approach to amur tiger re-identification, in: ICIP

    Cheng, X., Zhu, J., Zhang, N., Wang, Q., Zhao, Q., 2020. Detection features as attention (defat): A keypoint-free approach to amur tiger re-identification, in: ICIP

    2020

Show all 68 references
  1. [9]

    Compte, A., Yan, Y., Cortés, X., Escalera, S., Jacques-Junior, J.C.,

  2. [10]

    Expert Systems with Applications 293, 128466

    Housed pig identification and tracking for precision livestock farming. Expert Systems with Applications 293, 128466

  3. [11]

    Visual cues elicit differential ag- gression towards female and female mimics in the corkwing wrasse

    Ellis, B.A., Sørdalen, T.K., Briffa, M., Skiftesvik, A.B., Wilson, A.D.M., Halvorsen, K.T., 2026. Visual cues elicit differential ag- gression towards female and female mimics in the corkwing wrasse. Behavioral Ecology , arag022doi:10.1093/beheco/arag022

    2026 doi

  4. [12]

    Re-identification of fish individuals of undulate skate via deep learning within a few-shot context

    Gómez-Vargas, N., Alonso-Fernández, A., Blanquero, R., Antelo, L.T., 2023. Re-identification of fish individuals of undulate skate via deep learning within a few-shot context. Ecological Informatics 75, 102036

    2023

  5. [13]

    Delvingdeepintorectifiers: Surpassing human-level performance on imagenet classification, in: ProceedingsoftheIEEEinternationalconferenceoncomputervision, pp

    He,K.,Zhang,X.,Ren,S.,Sun,J.,2015. Delvingdeepintorectifiers: Surpassing human-level performance on imagenet classification, in: ProceedingsoftheIEEEinternationalconferenceoncomputervision, pp. 1026–1034

    2015

  6. [14]

    Animal re- identification algorithm for posture diversity, in: IEEE ICASSP

    He, Z., Qian, J., Yan, D., Wang, C., Xin, Y., 2023. Animal re- identification algorithm for posture diversity, in: IEEE ICASSP

    2023

  7. [15]

    Estimatingpopu- lation size, structure, and residency time for whale sharksrhincodon typusthroughcollaborativephoto-identification

    Holmberg,J.,Norman,B.,Arzoumanian,Z.,2009. Estimatingpopu- lation size, structure, and residency time for whale sharksrhincodon typusthroughcollaborativephoto-identification. EndangeredSpecies Research 7, 39–53. doi:10.3354/esr00186

    2009 doi

  8. [16]

    Parameter- efficient transfer learning for nlp, in: International conference on machine learning, PMLR

    Houlsby, N., Giurgiu, A., Jastrzebski, S., Morrone, B., De Larous- silhe, Q., Gesmundo, A., Attariyan, M., Gelly, S., 2019. Parameter- efficient transfer learning for nlp, in: International conference on machine learning, PMLR. pp. 2790–2799

    2019

  9. [17]

    Lora: Low-rank adaptation of large language models, in: ICLR

    Hu, E.J., Shen, Y., Wallis, P., Allen-Zhu, Z., Li, Y., Wang, S., Wang, L., Chen, W., 2022. Lora: Low-rank adaptation of large language models, in: ICLR

    2022

  10. [18]

    Unifor- mity and deformation: A benchmark for multi-fish real-time tracking in the farming

    Huang, J., Yu, X., An, D., Ning, X., Liu, J., Tiwari, P., 2025. Unifor- mity and deformation: A benchmark for multi-fish real-time tracking in the farming. Expert Systems with Applications 264, 125653

    2025

  11. [19]

    Visual prompt tuning, in: European conference on computer vision, Springer

    Jia,M.,Tang,L.,Chen,B.C.,Cardie,C.,Belongie,S.,Hariharan,B., Lim, S.N., 2022. Visual prompt tuning, in: European conference on computer vision, Springer. pp. 709–727

    2022

  12. [20]

    Toward re-identifying any animal

    Jiao,B.,Liu,L.,Gao,L.,Wu,R.,Lin,G.,Wang,P.,Zhang,Y.,2023. Toward re-identifying any animal. Neurips 36, 40042–40053

    2023

  13. [21]

    Body coloration as a dynamic signal during intrasexual communication in a cichlid fish

    John, L., Rick, I.P., Vitt, S., Thünken, T., 2021. Body coloration as a dynamic signal during intrasexual communication in a cichlid fish. BMC Zoology 6, 9

    2021

  14. [22]

    A survey of advances in vision-based vehicle re-identification

    Khan, S.D., Ullah, H., 2019. A survey of advances in vision-based vehicle re-identification. Computer Vision and Image Understanding 182, 50–63

    2019

  15. [23]

    Proxy anchor loss for deep metric learning, in: IEEE/CVF CVPR, pp

    Kim, S., Kim, D., Cho, M., Kwak, S., 2020. Proxy anchor loss for deep metric learning, in: IEEE/CVF CVPR, pp. 3238–3247

    2020

  16. [24]

    Transformer-basedsim- ilarity learning for re-identification of chickens

    Lamping,C.,Kootstra,G.,Derks,M.,2025. Transformer-basedsim- ilarity learning for re-identification of chickens. Smart Agricultural Technology

    2025

  17. [25]

    The power of scale for parameter-efficient prompt tuning, in: Proceedings of EMNLP

    Lester, B., Al-Rfou, R., Constant, N., 2021. The power of scale for parameter-efficient prompt tuning, in: Proceedings of EMNLP

    2021

  18. [26]

    A richly annotated pedestrian dataset for person retrieval in real surveillance scenarios

    Li, D., Zhang, Z., Chen, X., Huang, K., 2018. A richly annotated pedestrian dataset for person retrieval in real surveillance scenarios. IEEE transactions on image processing 28, 1575–1590

    2018

  19. [27]

    In- dividual dairy cow identification based on lightweight convolutional neural network

    Li, S., Fu, L., Sun, Y., Mu, Y., Chen, L., Li, J., Gong, H., 2021. In- dividual dairy cow identification based on lightweight convolutional neural network. Plos one 16, e0260510

    2021

  20. [28]

    Atrw: A benchmark for amur tiger re-identification in the wild, in: ACM MM

    Li, S., Li, J., Tang, H., Qian, R., Lin, W., 2020. Atrw: A benchmark for amur tiger re-identification in the wild, in: ACM MM

    2020

  21. [29]

    Clip-reid: exploiting vision-language model for image re-identification without concrete text labels, in: Proceedings of the AAAI conference on artificial intelligence, pp

    Li, S., Sun, L., Li, Q., 2023. Clip-reid: exploiting vision-language model for image re-identification without concrete text labels, in: Proceedings of the AAAI conference on artificial intelligence, pp. 1405–1413

    2023

  22. [30]

    Metawild: A multimodal dataset for animal re-identification with en- vironmentalmetadata,in:Proceedingsofthe33rdACMInternational Conference on Multimedia, pp

    Li, Y., Zhao, D., Qiao, T., Wu, Y., Pang, B., Koh, Y.S., 2025. Metawild: A multimodal dataset for animal re-identification with en- vironmentalmetadata,in:Proceedingsofthe33rdACMInternational Conference on Multimedia, pp. 13009–13015

    2025

  23. [31]

    Rede- formtr: Wildlife re-identification based on light-weight deformable transformer with multi-image feature fusion

    Li, Z., Yan, Z., Tian, W., Zeng, D., Liu, Y., Li, W., 2024. Rede- formtr: Wildlife re-identification based on light-weight deformable transformer with multi-image feature fusion. IEEE Access

    2024

  24. [32]

    Fishtrack: Multi- object tracking method for fish using spatiotemporal information fusion

    Liu, Y., Li, B., Zhou, X., Li, D., Duan, Q., 2024. Fishtrack: Multi- object tracking method for fish using spatiotemporal information fusion. Expert Systems with Applications 238, 122194

    2024

  25. [33]

    Novel use of deep neural networks on photographic identification of epaulette sharks across life stages

    Lonati, M., Jahanbakht, M., Atkins, D., Bierwagen, S.L., Chin, A., Barnett, A., Rummer, J.L., 2024. Novel use of deep neural networks on photographic identification of epaulette sharks across life stages. Journal of Fish Biology 105, 1572–1587. 23

    2024

  26. [34]

    Decoupled weight decay regulariza- tion

    Loshchilov, I., Hutter, F., 2017. Decoupled weight decay regulariza- tion. arXiv preprint arXiv:1711.05101

    2017 arXiv

  27. [35]

    Mark-resight abundance estimation under incompleteidentificationofmarkedindividuals

    McClintock, B.T., Hill, J.M., Fritz, L., Chumbley, K., Luxa, K., Diefenbach, D.R., 2014. Mark-resight abundance estimation under incompleteidentificationofmarkedindividuals. MethodsinEcology and Evolution 5, 1294–1304

    2014

  28. [36]

    Robust re-identification of manta rays from natural markings by learning pose invariant embeddings, in: DICTA

    Moskvyak, O., Maire, F., Dayoub, F., Armstrong, A., Baktashmot- lagh, M., 2021. Robust re-identification of manta rays from natural markings by learning pose invariant embeddings, in: DICTA

    2021

  29. [37]

    Learning landmark guided embeddings for animal re-identification, in: IEEE WACVw

    Moskvyak, O., Maire, F., Dayoub, F., Baktashmotlagh, M., 2020. Learning landmark guided embeddings for animal re-identification, in: IEEE WACVw

    2020

  30. [38]

    Murali, N., Schneider, J., Levine, J., Taylor, G., 2019. Classifica- tion and re-identification of fruit fly individuals across days with convolutional neural networks, in: 2019 IEEE Winter Conference on Applications of Computer Vision (WACV), IEEE. pp. 570–578

    2019

  31. [39]

    Species-agnostic patterned animal re-identification by aggregating deep local features

    Nepovinnykh,E.,Chelak,I.,Eerola,T.,Immonen,V.,Kälviäinen,H., Kholiavchenko, M., Stewart, C., 2024a. Species-agnostic patterned animal re-identification by aggregating deep local features. IJCV

  32. [40]

    Sealid: Saimaa ringed seal re- identification dataset

    Nepovinnykh, E., Eerola, T., Biard, V., Mutka, P., Niemi, M., Kun- nasranta, M., Kälviäinen, H., 2022. Sealid: Saimaa ringed seal re- identification dataset. Sensors

    2022

  33. [41]

    Siamese network based pelage pattern matching for ringed seal re-identification, in: IEEE WACVw

    Nepovinnykh, E., Eerola, T., Kalviainen, H., 2020. Siamese network based pelage pattern matching for ringed seal re-identification, in: IEEE WACVw

    2020

  34. [42]

    Norppa: Novel ringed seal re-identification by pelage pattern aggre- gation, in: IEEE WACVw

    Nepovinnykh, E., Eerola, T., Kalviainen, H., Chelak, I., 2024b. Norppa: Novel ringed seal re-identification by pelage pattern aggre- gation, in: IEEE WACVw

  35. [43]

    Re-identification of patterned animals by multi-image feature aggregation and geometric similarity

    Nepovinnykh, E., Immonen, V., Eerola, T., Stewart, C., Kälviäinen, H., 2025. Re-identification of patterned animals by multi-image feature aggregation and geometric similarity. IET Computer Vision

    2025

  36. [44]

    A contrastive learning approach for individual re-identification in a wild fish population, in: Proceedings of the Northern Lights Deep Learning Workshop

    Olsen, Ø.L., Sørdalen, T.K., Goodwin, M., Malde, K., Knausgård, K.M., Halvorsen, K.T., 2023. A contrastive learning approach for individual re-identification in a wild fish population, in: Proceedings of the Northern Lights Deep Learning Workshop

    2023

  37. [45]

    Film:Visualreasoningwithageneralconditioninglayer,in:Proceed- ings of the AAAI Conference on Artificial Intelligence

    Perez, E., Strub, F., de Vries, H., Dumoulin, V., Courville, A., 2018. Film:Visualreasoningwithageneralconditioninglayer,in:Proceed- ings of the AAAI Conference on Artificial Intelligence

    2018

  38. [46]

    Dynamic multi-behaviour, orientation-invariant re-identification of holstein- friesian cattle

    Perneel,M.,Adriaens,I.,Verwaeren,J.,Aernouts,B.,2025. Dynamic multi-behaviour, orientation-invariant re-identification of holstein- friesian cattle. Sensors

    2025

  39. [47]

    Pigments, patterns, and fish behavior

    Price, A.C., Weadick, C.J., Shim, J., Rodd, F.H., 2008. Pigments, patterns, and fish behavior. Zebrafish 5, 297–307

    2008

  40. [48]

    Auto-reid: Searching for a part-aware convnet for person re-identification, in: ProceedingsoftheIEEE/CVFInternationalConferenceonComputer Vision, pp

    Quan, R., Dong, X., Wu, Y., Zhu, L., Yang, Y., 2019. Auto-reid: Searching for a part-aware convnet for person re-identification, in: ProceedingsoftheIEEE/CVFInternationalConferenceonComputer Vision, pp. 3750–3759

    2019

  41. [49]

    8748–8763

    Radford,A.,Kim,J.W.,Hallacy,C.,etal.,2021.Learningtransferable visual models from natural language supervision, in: ICML, pp. 8748–8763

    2021

  42. [50]

    Sam 2: Segment anything in images and videos

    Ravi, N., Gabeur, V., Hu, Y.T., Hu, R., Ryali, C., Ma, T., Khedr, H., Rädle, R., Rolland, C., Gustafson, L., et al., 2024. Sam 2: Segment anything in images and videos. arXiv preprint arXiv:2408.00714

    2024 arXiv

  43. [51]

    Deep learning methods for multi-species animal re-identification and tracking: a survey

    Ravoor, P., T.s.b., S., 2020. Deep learning methods for multi-species animal re-identification and tracking: a survey. Computer Science Review

    2020

  44. [52]

    Individualidentificationofbrownbearsusingpose- aware metric learning

    Rosenberg, B., Zhou, M., Wolf, N., Mathis, M.W., Harris, B.P., Mathis,A.,2026. Individualidentificationofbrownbearsusingpose- aware metric learning. Current Biology 36, 645–659.e14. doi:https: //doi.org/10.1016/j.cub.2025.12.022

    2026 doi

  45. [53]

    A wild fish image dataset for individual re-identification and phenotyping

    Sordalen,T.K.,Malde,K.,Skiftesvik,A.B.,Sauvaitre,C.,Beyan,C., Larsen, T.and Halvorsen, K.T., 2026. A wild fish image dataset for individual re-identification and phenotyping. Scientific data

    2026

  46. [54]

    Melopsreid:Awildfishimage dataset for re-identification

    Sørdalen,T.K.,Halvorsen,K.T.,2025. Melopsreid:Awildfishimage dataset for re-identification. URL:https://doi.org/10.5281/zenodo. 17099925, doi:10.5281/zenodo.17099925

    2025 doi

  47. [55]

    Manta matcher: Automated photographic identification of manta rays using keypoint features

    Town, C., Marshall, A., Sethasathien, N., 2013. Manta matcher: Automated photographic identification of manta rays using keypoint features. EcologyandEvolution3,1902–1914. doi:10.1002/ece3.587

    2013 doi

  48. [56]

    Phenotypic vari- ation between dimorphic males in corkwing wrasse

    Uglem, I., Rosenqvist, G., Wasslavik, H.S., 2000. Phenotypic vari- ation between dimorphic males in corkwing wrasse. Journal of Fish Biology 57, 1–14

    2000

  49. [57]

    Animal re- identification in video through track clustering

    Williams, F., Hennessey, S., Kuncheva, L., 2025. Animal re- identification in video through track clustering. PAA

    2025

  50. [58]

    Deep features for person re-identification on metric learning

    Wu, W., Tao, D., Li, H., Yang, Z., Cheng, J., 2021. Deep features for person re-identification on metric learning. Pattern Recognition 110

    2021

  51. [59]

    An individual identity-drivenframeworkforanimalre-identification

    Wu, Y., Zhao, D., Zhang, J., Koh, Y.S., 2024. An individual identity-drivenframeworkforanimalre-identification. arXivpreprint arXiv:2410.22927

    2024 arXiv

  52. [60]

    Holstein- friesian re-identification using multiple cameras and self-supervision on a working farm

    Yu, P., Burghardt, T., Dowsey, A., Campbell, N., 2025. Holstein- friesian re-identification using multiple cameras and self-supervision on a working farm. Computers and Electronics in Agriculture

    2025

  53. [61]

    Zaken, E.B., Goldberg, Y., Ravfogel, S., 2022. Bitfit: Sim- ple parameter-efficient fine-tuning for transformer-based masked language-models, in: Proceedings of the 60th Annual Meeting of the AssociationforComputationalLinguistics(Volume2:ShortPapers), pp. 1–9

    2022

  54. [62]

    Expert Systems with Applications , 128541

    Zhang,K.,Qin,Y.,Pan,H.,Ma,B.,2025.Diffusion-basedadversarial attack method against person re-identification. Expert Systems with Applications , 128541

    2025

  55. [63]

    Yakreid-103: A benchmark for yak re-identification, in: IEEE IJCB

    Zhang, T., Zhao, Q., Da, C., Zhou, L., Li, L., Jiancuo, S., 2021. Yakreid-103: A benchmark for yak re-identification, in: IEEE IJCB

    2021

  56. [64]

    Wild terrestrial animal re- identification based on an improved locally aware transformer with a cross-attention mechanism

    Zheng, Z., Zhao, Y., Li, A., Yu, Q., 2022. Wild terrestrial animal re- identification based on an improved locally aware transformer with a cross-attention mechanism. Animals

    2022

  57. [65]

    Learning to prompt for vision-language models, in: CVPR

    Zhou, K., Yang, J., Loy, C.C., Liu, Z., 2022. Learning to prompt for vision-language models, in: CVPR

    2022

  58. [66]

    Voc-reid: Vehicle re- identificationbasedonvehicle-orientation-camera,in:Proceedingsof theIEEE/CVFConferenceonComputerVisionandPatternRecogni- tion Workshops, pp

    Zhu, X., Luo, Z., Fu, P., Ji, X., 2020. Voc-reid: Vehicle re- identificationbasedonvehicle-orientation-camera,in:Proceedingsof theIEEE/CVFConferenceonComputerVisionandPatternRecogni- tion Workshops, pp. 602–603

    2020

  59. [67]

    Polarbearvidid: A video-based re-identification benchmark dataset for polar bears

    Zuerl, M., Dirauf, R., Koeferl, F., Steinlein, N., Sueskind, J., Zanca, D., Brehm, I., Fersen, L.v., Eskofier, B., 2023. Polarbearvidid: A video-based re-identification benchmark dataset for polar bears. Animals 13, 801

    2023

  60. [68]

    Wildlife- datasets:Anopen-sourcetoolkitforanimalre-identification,in:IEEE WACV

    Čermák, V., Picek, L., Adam, L., Papafitsoros, K., 2024. Wildlife- datasets:Anopen-sourcetoolkitforanimalre-identification,in:IEEE WACV. 24

    2024

Pith tools

Reviewed July 13, 2026 · model on record in the stance chip above.