Pith. sign in

REVIEW 4 major objections 7 minor 53 references

Cross-Domain Few-Shot Learning via Multi-View Collaborative Optimization with Vision-Language Models

T0 review · 4 major / 7 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 The paper proposes CoMuCo, a two-expert fine-tuning strategy with Laplace and Jeffreys regularizers, and claims it consistently beats existing prompt- and adapter-tuning methods on few-shot benchmarks, especially under cross-domain shift.

desk verdict Useful fine-tuning recipe and a good new benchmark, but the multi-view story is untested and an ablation row looks off. read the letter →

arxiv 2508.12861 v1 pith:AKX7VO75 submitted 2025-08-18 cs.CV

Dexia Chen , Wentao Zhang , Qianjie Zhu , Ping Hu , Weibing Li , Tong Zhang , Ruixuan Wang This is my paper · ORCID
classification cs.CV
keywords few-shotlearningcross-domainvision-languagemodelsCLIPefficienttransferJeffreysdivergenceLaplacepriormulti-view
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

CoMuCo is a fine-tuning strategy for vision-language models aimed at few-shot classification when the target domain is far from natural images—skin lesions, pathology slides, retinal fundus images, remote sensing scenes, steel defects, crop pests, and galaxy images. It trains two experts from the same CLIP checkpoint: a Feature Integrator that changes only the final attention-pooling block, and a Feature Refiner that tunes deeper layers, so the two views favor stability and adaptation respectively. Two regularizers steer the training: an L1 penalty on each expert's logit deviation from frozen zero-shot CLIP (equivalent to a Laplace prior) that preserves general knowledge, and a Jeffreys-divergence consensus term that pulls the two experts' prediction distributions together. On both the established CLIP benchmark and a new seven-dataset cross-domain benchmark, the paper reports consistent gains over existing prompt- and adapter-tuning methods, with margins that grow as the number of training images per class increases.

What carries the argument

The load-bearing objects are the two expert modules—Feature Integrator (FI) and Feature Refiner (FR)—and the two regularization terms that coordinate them. FI is the invariant expert, updated only at the final attention-pooling block; FR is the adaptive expert, fine-tuned in deeper layers. The prior consistency constraint models each expert's logit deviation from frozen zero-shot CLIP as a zero-mean Laplace distribution, which is equivalent to $\ell^1$ regularization on the logit offset, a mechanism that permits sparse class-level changes without erasing general knowledge. The multi-view consensus constraint uses Jeffreys divergence, the symmetric KL divergence $D_J(p\parallel q)=D_{KL}(p\parallel q)+D_{KL}(q\parallel p)$, as a fourth-order approximation to the squared geodesic distance between the two experts' prediction distributions on the statistical manifold; minimizing it aligns the two views. These pieces work together: FI supplies stability, FR supplies adaptation, the Laplace term prevents forgetting, and the Jeffreys term makes the two views compatible, with final logits fused as $s_i=\alpha s_{FR}(x_i)+\beta s_{FI}(x_i)+\gamma s_{ZS}(x_i)$.

What would settle it

Train CoMuCo on a cross-domain dataset such as Galaxy10 DECaLS at 16-shot, then measure the Jeffreys divergence between the FR and FI prediction distributions on the test set; if the divergence is near zero on almost every sample, the two views have collapsed and the consensus term is not doing the work. A second decisive check is to ablate the fusion weights: if setting $\alpha=\beta=0$ (using only zero-shot CLIP logits) and training FI/FR separately reproduces most of the accuracy gains, the multi-view mechanism is not responsible.

Watch

Extended reading notes

Core claim

The paper's central claim is that the two-branch architecture, plus the two logit-space regularizers, is what lets CLIP transfer to genuinely different imaging domains. The Feature Integrator is the conservative expert: it keeps most of the pre-trained model intact and refines only attention pooling at the last block, preserving knowledge that is already relevant. The Feature Refiner is the adaptive expert: it tunes deeper layers to learn task-specific patterns the pre-training never saw. The final prediction is a weighted fusion of the two experts' logits and the frozen zero-shot CLIP logits ($\alpha=0.2$, $\beta=0.2$, $\gamma=0.6$). The prior consistency term penalizes the $\ell^1$ norm of each expert's logit offset from zero-shot CLIP, enforcing sparse changes; the consensus term minimizes the Jeffreys divergence between the experts' predictive distributions, which the paper proves approximates the squared geodesic distance on the statistical manifold to fourth order. Empirically, on the new cross-domain benchmark with ResNet-50, the reported average improvements over the best baseline are 0.9, 2.25, 2.78, 5.27, and 7.04 points at 1, 2, 4, 8, and 16 shots, with analogous gains under ViT-B/16 and on the standard CLIP benchmark.

Load-bearing premise

The load-bearing premise is that the Feature Integrator and Feature Refiner, both initialized from the same CLIP weights and both pulled toward the zero-shot CLIP logits, learn genuinely different and complementary features; if the two branches collapse into near-identical representations, the Jeffreys consensus term becomes vacuous and the reported gains would reduce to the fixed logit averaging with frozen CLIP.

Editorial extensions

If this is right

  • Efficient CLIP transfer to distant domains does not require large prompt or adapter modules; two modestly tuned branches plus two logit-space regularizers are enough to improve on existing methods.
  • The gains are not tied to one visual backbone: the paper reports consistent improvements with ResNet-50, ResNet-101, ViT-B/32, ViT-B/16, and ConvNeXt-Base, so the mechanism transfers across architectures.
  • The new seven-dataset cross-domain benchmark gives the community a common testbed where zero-shot CLIP is weak, allowing future few-shot methods to be compared under genuine domain shift.
  • Because the reported margins on the cross-domain benchmark widen with sample size (from 0.9 to 7.04 points on ResNet-50), the method is most valuable when a modest number of images is available, not only in the extreme 1-shot regime.
  • The domain-adaptation results (ImageNet trained, ImageNet-V2 and ImageNet-Sketch evaluated) indicate that the features learned by the two experts remain general enough to transfer beyond the training domain.

Reading between the lines

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

  • If the two experts are truly complementary, a testable extension is to replace the Jeffreys consensus term with another symmetric divergence (for example, the Bhattacharyya distance) and compare; similar gains would suggest the specific geometry matters less than the presence of a two-view agreement constraint.
  • A further extension is to initialize FI and FR from a prompt-tuned or adapter-tuned CLIP rather than from the frozen checkpoint, to see whether the collaborative mechanism stacks with existing transfer methods instead of only replacing them.
  • The paper's own tiering of the benchmark by zero-shot CLIP accuracy implies a prediction the authors leave implicit: gains should be largest on high-shift datasets such as the pathology and fundus sets, and smaller on low-shift sets such as remote sensing; readers can check this pattern in the reported per-dataset numbers.
  • Because the consensus term is only meaningful if the two branches diverge, a direct check of FI/FR representation similarity after training would settle whether the reported gains come from collaboration or from the fixed 0.2/0.2/0.6 fusion with zero-shot CLIP.
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

4 major / 7 minor

Summary. The paper proposes CoMuCo, a fine-tuning strategy for CLIP in few-shot settings, especially under cross-domain shifts. The method has two expert modules, a Feature Integrator (FI) trained shallowly and a Feature Refiner (FR) trained more deeply, whose logits are fused with frozen zero-shot CLIP logits. Two regularizers are added: an L1-style prior consistency loss that keeps each branch's logits close to CLIP zero-shot logits, and a Jeffreys-divergence consensus loss that aligns the two branches' predictive distributions. The paper also introduces a new cross-domain few-shot benchmark of seven datasets and reports experiments on this benchmark and on the CLIP Benchmark with ResNet-50 and ViT-B/16, claiming consistent state-of-the-art performance.

Significance. If the empirical claims hold, CoMuCo is a simple and plausible approach for transferring CLIP to domains that differ from its natural-image pretraining distribution. The new cross-domain benchmark addresses a real gap in few-shot CLIP evaluation, and the proposed regularizers are easy to implement. The theoretical components are standard: the Laplace/L1 equivalence is well known, and the Jeffreys divergence is a standard symmetric KL divergence; the geometric framing is motivation rather than a load-bearing derivation. The paper plans to release code and the benchmark, which would aid reproducibility. However, the empirical support as presented is incomplete: no variance estimates are reported, the central multi-view complementarity premise is not quantitatively verified, several referenced tables and proofs are missing, and some headline numbers conflict with the tables. These issues currently limit the significance of the claimed results.

major comments (4)
  1. [Experiments, Implementation; Tables 6, 7, 9] All results are stated to be averaged over three seeds, but no standard deviation, confidence interval, or per-seed result is reported anywhere in Tables 6, 7, 9 or Figures 4 and 5. Several headline gains are small; for example, the ResNet-50 1-shot cross-domain average improvement is 0.9 percentage points (Table 6: 38.19 vs. 37.29), and on individual 1-shot datasets CoMuCo is worse than the best baseline (Table 6: Skin40 27.57 vs. TCP 29.40; TCGA12 18.23 vs. TCP 24.57). Without variance estimates or significance tests, the claim that CoMuCo 'consistently outperforms current methods' is not statistically supported. Please report standard deviations over the three seeds and, ideally, per-dataset win/loss counts.
  2. [Eq. (5), Eq. (9), Dual-Expert Framework, Appendix C.5] The central mechanism is multi-view collaboration, but the manuscript does not establish that FI and FR actually provide complementary views. Both modules are initialized from the same CLIP weights and are trained with the same cross-entropy objective on the same data; the only architectural difference is fine-tuning depth. The consensus loss in Eq. (5) minimizes the Jeffreys divergence between the two branches, and the total objective in Eq. (9) contains no term that rewards diversity, so the branches could collapse to near-duplicate predictors, making the consensus term vacuous. The only evidence offered in Appendix D is qualitative (GradCAM and t-SNE). Please report a quantitative diversity measure, such as the average cosine similarity between FI and FR features or logits before and after training, or the converged value of the Jeffreys term, and include the ensemble control from Appendix C.5 (Table 11), which is currently missing. Without this, the reported gains over the fixed 0.2/0.2/0.6 logit averaging with zero-shot CLIP cannot be attributed to the proposed collaborative mechanism.
  3. [Efficacy of the Proposed Method, Appendix C.3; Tables 7 and 9] The numerical claims in the main text are not consistent with the reported tables. For ViT-B/16 on the cross-domain benchmark, the text reports improvements of 2.03%, 3.23%, 3.59%, 2.83%, and 2.78% over the best baseline for 1/2/4/8/16-shot, but Table 7 yields best-baseline margins of 2.26, 2.72, 2.43, 2.47, and 1.75 percentage points. For ResNet-50 on the CLIP Benchmark, the main text reports gains of 1.48%, 2.77%, and 4.65% for 1/4/16-shot, while Appendix C.3 reports 1.55%, 2.82%, and 4.67% for the same settings. The authors should reconcile these numbers and state exactly which baseline and which table each margin is computed against.
  4. [Appendix B.1, Table 3, Appendix C.2] The new cross-domain benchmark is self-constructed and currently lacks external validation, and it contains internal inconsistencies that block reproduction. The RFMiD dataset is called 'RFMiD12' in Tables 5-7 and the main text, but Appendix B.1 says the benchmark uses a subset of the 15 most populous classes and Table 3 lists 15 classes; the relation between the name 'RFMiD12' and the 15-class subset is unclear. The exact train/validation protocol is also not specified for every dataset (e.g., TCGA12 and Skin40 are listed without a validation split, while IP102 uses its official split). Please clarify the class counts and splits, and provide the exact sampling procedure or image indices to enable external checks.
minor comments (7)
  1. [Table 1] The first row of Table 1 has no components checked and its values (58.18, 55.61, 13.90) coincide exactly with the zero-shot CLIP numbers in Tables 9 and 6. If this row is the zero-shot CLIP baseline, label it as such; if it is an LCE-only trained model, explain why training with 16-shot cross-entropy leaves the accuracy unchanged to two decimals.
  2. [Impact of Fine-tuning Layer Configurations in FR] The text contains an unresolved cross-reference '?? shows' when discussing the fine-tuning depth study; this should point to the relevant figure (apparently Fig. 9).
  3. [Theorem 1 and Theorem 2] Theorem 1 and Theorem 2 refer to proofs in Appendix A.1 and A.2, but the appendices included in the manuscript start at B; the proofs should be included or the references corrected.
  4. [Appendix C.3 and C.5] Appendix C.3 states that results are presented in 'Tab. 9 and Tab. 10', and Appendix C.5 refers to 'Tab. 11'; Tables 10 and 11 do not appear in the manuscript. Please include them.
  5. [Abstract and Conclusion] The abstract and conclusion say CoMuCo 'consistently outperforms current methods', but Tables 6 and 7 show several per-dataset cases where it is below the best baseline (e.g., Table 6, 1-shot Skin40 and TCGA12; Table 7, 1-shot IP102). Please qualify the claim as average performance or report per-dataset win/loss counts.
  6. [Eq. (7)] In Eq. (7), the coefficients alpha, beta, and gamma are introduced without constraints; the text later fixes alpha = beta = 0.2, implying gamma = 0.6, but it would be clearer to state the simplex constraint gamma = 1 - alpha - beta explicitly.
  7. [Figures 4 and 5] Figures 4 and 5 show a legend that omits LP++, although LP++ results appear in Tables 6, 7, and 9; please either add it to the figures or explain why it is omitted.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: CoMuCo's losses, fusion rule, and ablations are self-contained; benchmark gains are direct comparisons with fixed hyperparameters.

full rationale

The paper's derivation chain is self-contained and no prediction reduces to a fitted value by the paper's own equations. The training objective (Eq. 9) is explicit: cross-entropy on the fused logits (Eq. 7) plus L1 prior losses (Eq. 3) and Jeffreys divergence consensus (Eq. 5), with fixed coefficients alpha=beta=0.2 and lambda1=lambda2=lambda3=0.1, and a sensitivity study is reported. The Laplace/L1 equivalence (Theorem 1) and Jeffreys/geodesic approximation (Theorem 2) are standard mathematical facts, not functions of the method's outputs. The central claims are empirical comparisons on established and newly assembled benchmarks; no parameter is fitted to the test results and then renamed a prediction. The only self-citation found is the Skin40 dataset source (Yang et al. 2023, which shares co-author R. Wang), but this is an externally available dataset, not a load-bearing theoretical premise, so it does not raise the circularity score. Two completeness issues are noted but are not circularity: Appendix C.5 references 'Tab. 11' for the ensemble control, and the layer-configuration subsection contains a dangling '?? shows that'; both tables/figures are absent from the provided text. Likewise, the untested complementarity of FI and FR (both initialized from CLIP and aligned by the consensus term) is a correctness/verification risk, not a circular reduction: the paper does not define the gains in terms of that premise.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The method introduces no new particles, forces, or mathematical entities. The free parameters are five hand-set scalar weights, and the theoretical framework reuses standard Laplace/L1 and Jeffreys-geodesic results. The load-bearing assumptions are the complementarity of the two expert branches and the representativeness of the new benchmark.

free parameters (6)
  • Fusion weight alpha for FR logits = 0.2
    Coefficient of s_FR in the fused logits (Eq. 7). Fixed by hand across all datasets.
  • Fusion weight beta for FI logits = 0.2
    Coefficient of s_FI in the fused logits (Eq. 7). Fixed by hand; gamma is derived as 0.6.
  • Prior consistency weight lambda1 = 0.1
    Weight for L_R = ||s_FR - s_ZS||_1 in the total objective (Eq. 9).
  • Prior consistency weight lambda2 = 0.1
    Weight for L_I = ||s_FI - s_ZS||_1 in the total objective (Eq. 9).
  • Consensus weight lambda3 = 0.1
    Weight for the Jeffreys divergence L_D in the total objective (Eq. 9).
  • Laplace scale b = absorbed into lambda1 and lambda2
    The scale b in Definition 1 cancels in Theorem 1 and is effectively absorbed by the lambda weights; no independent value is needed.
assumptions (5)
  • domain assumption CLIP's frozen representations plus modest fine-tuning transfer to downstream tasks
    The whole method presupposes that small trainable edits to CLIP features improve few-shot recognition; this is the shared assumption of the cited CoOp, Tip-Adapter, TaskRes, and related works.
  • standard math Jeffreys divergence approximates the squared Fisher-Rao geodesic distance to fourth order
    Theorem 2, with proof in Appendix A.2 (not visible in the provided text). This is a standard result in information geometry and is used to motivate the consensus loss.
  • ad hoc to paper Components of the logit deviation vector are i.i.d. zero-mean Laplace
    Definition 1 is a modeling choice that turns Theorem 1 into an L1 penalty; it is not derived from data or from the structure of CLIP.
  • domain assumption The two expert branches initialized from the same CLIP weights learn complementary views
    The dual-expert design claims functional complementarity (Dual-Expert Framework section). If this fails, the consensus term is redundant and the gains would come only from logit averaging.
  • domain assumption The newly constructed cross-domain benchmark is representative of relevant cross-domain imaging tasks
    The benchmark is assembled by the authors from seven public datasets with their own subsetting and split rules; the conclusions about cross-domain performance depend on this benchmark being fair and representative.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Cross-Domain Few-Shot Learning via Multi-View Collaborative Optimization with Vision-Language Models." pith.science (2026). https://pith.science/paper/AKX7VO75

@misc{pith2026250812861,
  author       = {Pith},
  title        = {Pith review of: Cross-Domain Few-Shot Learning via Multi-View Collaborative Optimization with Vision-Language Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AKX7VO75}},
  note         = {Machine review of arXiv:2508.12861}
}
read the original abstract

Vision-language models (VLMs) pre-trained on natural image and language data, such as CLIP, have exhibited significant potential in few-shot image recognition tasks, leading to development of various efficient transfer learning methods. These methods exploit inherent pre-learned knowledge in VLMs and have achieved strong performance on standard image datasets. However, their effectiveness is often limited when confronted with cross-domain tasks where imaging domains differ from natural images. To address this limitation, we propose Consistency-guided Multi-view Collaborative Optimization (CoMuCo), a novel fine-tuning strategy for VLMs. This strategy employs two functionally complementary expert modules to extract multi-view features, while incorporating prior knowledge-based consistency constraints and information geometry-based consensus mechanisms to enhance the robustness of feature learning. Additionally, a new cross-domain few-shot benchmark is established to help comprehensively evaluate methods on imaging domains distinct from natural images. Extensive empirical evaluations on both existing and newly proposed benchmarks suggest CoMuCo consistently outperforms current methods in few-shot tasks. The code and benchmark will be released.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 37 canonical work pages

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Allen - Zhu, Z.; and Li, Y. 2023. Towards Understanding Ensemble, Knowledge Distillation and Self-Distillation in Deep Learning. In ICLR

    work page 2023

  4. [4]

    Bossard, L.; Guillaumin, M.; and Gool, L. V. 2014. Food-101 - Mining Discriminative Components with Random Forests. In ECCV

    work page 2014

  5. [5]

    Chen, G.; Yao, W.; Song, X.; Li, X.; Rao, Y.; and Zhang, K. 2023. PLOT: Prompt Learning with Optimal Transport for Vision-Language Models. In ICLR

    work page 2023

  6. [6]

    J.; Chen, C.; Li, Y.; Chen, T

    Chen, R. J.; Chen, C.; Li, Y.; Chen, T. Y.; Trister, A. D.; Krishnan, R. G.; and Mahmood, F. 2022. Scaling vision transformers to gigapixel images via hierarchical self-supervised learning. In CVPR

    work page 2022

  7. [7]

    Cheng, G.; Han, J.; and Lu, X. 2017. Remote sensing image scene classification: Benchmark and state of the art. Proceedings of the IEEE, 105(10): 1865--1883

    2017

  8. [8]

    Cherti, M.; Beaumont, R.; Wightman, R.; Wortsman, M.; Ilharco, G.; Gordon, C.; Schuhmann, C.; Schmidt, L.; and Jitsev, J. 2023. Reproducible scaling laws for contrastive language-image learning. In CVPR, 2818--2829

    2023

Show all 53 references
  1. [9]

    Cimpoi, M.; Maji, S.; Kokkinos, I.; Mohamed, S.; and Vedaldi, A. 2014. Describing Textures in the Wild. In CVPR

    2014

  2. [10]

    Deng, J.; Dong, W.; Socher, R.; Li, L.; Li, K.; and Fei - Fei, L. 2009. ImageNet: A large-scale hierarchical image database. In CVPR

    2009

  3. [11]

    Fei - Fei, L.; Fergus, R.; and Perona, P. 2007. Learning generative visual models from few training examples: An incremental Bayesian approach tested on 101 object categories. Comput. Vis. Image Underst., 106(1): 59--70

    2007

  4. [12]

    Gao, P.; Geng, S.; Zhang, R.; Ma, T.; Fang, R.; Zhang, Y.; Li, H.; and Qiao, Y. 2024 a . Clip-adapter: Better vision-language models with feature adapters. International Journal of Computer Vision, 132(2): 581--595

    2024

  5. [13]

    Gao, Y.; Liu, J.; Xu, Z.; Wu, T.; Zhang, E.; Li, K.; Yang, J.; Liu, W.; and Sun, X. 2024 b . Softclip: Softer cross-modal alignment makes clip stronger. In AAAI

    2024

  6. [14]

    Gharoun, H.; Momenifar, F.; Chen, F.; and Gandomi, A. 2024. Meta-learning approaches for few-shot learning: A survey of recent advances. ACM Computing Surveys, 56(12): 1--41

    2024

  7. [15]

    Guo, Z.; Zhang, R.; Qiu, L.; Ma, X.; Miao, X.; He, X.; and Cui, B. 2023. CALIP: Zero-Shot Enhancement of CLIP with Parameter-Free Attention. In AAAI

    2023

  8. [16]

    Helber, P.; Bischke, B.; Dengel, A.; and Borth, D. 2019. EuroSAT : A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classification. IEEE J. Sel. Top. Appl. Earth Obs. Remote. Sens. , 12(7): 2217--2226

    2019

  9. [17]

    Huang, Y.; Shakeri, F.; Dolz, J.; Boudiaf, M.; Bahig, H.; and Ben Ayed, I. 2024 a . LP++: A Surprisingly Strong Linear Probe for Few-Shot CLIP. In CVPR

    2024

  10. [18]

    Huang, Y.; Tang, J.; Chen, Z.; Zhang, R.; Zhang, X.; Chen, W.; Zhao, Z.; Zhao, Z.; Lv, T.; Hu, Z.; and Zhang, W. 2024 b . Structure-CLIP: Towards Scene Graph Knowledge to Enhance Multi-modal Structured Representations. In AAAI

    2024

  11. [19]

    Jia, C.; Yang, Y.; Xia, Y.; Chen, Y.-T.; Parekh, Z.; Pham, H.; Le, Q.; Sung, Y.-H.; Li, Z.; and Duerig, T. 2021. Scaling up visual and vision-language representation learning with noisy text supervision. In ICML

    2021

  12. [20]

    U.; Rasheed, H

    Khattak, M. U.; Rasheed, H. A.; Maaz, M.; Khan, S. H.; and Khan, F. S. 2023. MaPLe: Multi-modal Prompt Learning. In CVPR

    2023

  13. [21]

    Krause, J.; Stark, M.; Deng, J.; and Fei - Fei, L. 2013. 3D Object Representations for Fine-Grained Categorization. In ICCV

    2013

  14. [22]

    W.; and Bovy, J

    Leung, H. W.; and Bovy, J. 2019. Deep learning of multi-element abundances from high-resolution spectroscopic data. Monthly Notices of the Royal Astronomical Society, 483(3): 3255--3277

    2019

  15. [23]

    Li, Y.; Fan, H.; Hu, R.; Feichtenhofer, C.; and He, K. 2023. Scaling language-image pre-training via masking. In CVPR

    2023

  16. [24]

    B.; and Vedaldi, A

    Maji, S.; Rahtu, E.; Kannala, J.; Blaschko, M. B.; and Vedaldi, A. 2013. Fine-Grained Visual Classification of Aircraft. CoRR, abs/1306.5151

    2013 arXiv

  17. [25]

    Nilsback, M.; and Zisserman, A. 2008. Automated Flower Classification over a Large Number of Classes. In Sixth Indian Conference on Computer Vision, Graphics & Image Processing, ICVGIP 2008, Bhubaneswar, India, 16-19 December 2008 , 722--729

    2008

  18. [26]

    Panchal, S.; Naik, A.; Kokare, M.; Pachade, S.; Naigaonkar, R.; Phadnis, P.; and Bhange, A. 2023. Retinal Fundus Multi-Disease Image Dataset (RFMiD) 2.0: a dataset of frequently and rarely identified diseases. Data, 8(2): 29

    2023

  19. [27]

    M.; Vedaldi, A.; Zisserman, A.; and Jawahar, C

    Parkhi, O. M.; Vedaldi, A.; Zisserman, A.; and Jawahar, C. V. 2012. Cats and dogs. In CVPR

    2012

  20. [28]

    W.; Hallacy, C.; Ramesh, A.; Goh, G.; Agarwal, S.; Sastry, G.; Askell, A.; Mishkin, P.; Clark, J.; Krueger, G.; and Sutskever, I

    Radford, A.; Kim, J. W.; Hallacy, C.; Ramesh, A.; Goh, G.; Agarwal, S.; Sastry, G.; Askell, A.; Mishkin, P.; Clark, J.; Krueger, G.; and Sutskever, I. 2021. Learning Transferable Visual Models From Natural Language Supervision. In ICML

    2021

  21. [29]

    Recht, B.; Roelofs, R.; Schmidt, L.; and Shankar, V. 2019. Do ImageNet Classifiers Generalize to ImageNet? In ICML

    2019

  22. [30]

    Schuhmann, C.; Vencu, R.; Beaumont, R.; Kaczmarczyk, R.; Mullis, C.; Katta, A.; Coombes, T.; Jitsev, J.; and Komatsuzaki, A. 2021. LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs. CoRR, abs/2111.02114

    2021 arXiv

  23. [31]

    R.; Cogswell, M.; Das, A.; Vedantam, R.; Parikh, D.; and Batra, D

    Selvaraju, R. R.; Cogswell, M.; Das, A.; Vedantam, R.; Parikh, D.; and Batra, D. 2017. Grad-cam: Visual explanations from deep networks via gradient-based localization. In ICCV, 618--626

    2017

  24. [32]

    Song, K.; and Yan, Y. 2013. A noise robust method based on completed local binary patterns for hot-rolled steel strip surface defects. Applied Surface Science, 285: 858--864

    2013

  25. [33]

    R.; and Shah, M

    Soomro, K.; Zamir, A. R.; and Shah, M. 2012. UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild. CoRR, abs/1212.0402

    2012 arXiv

  26. [34]

    Y.; Wang, X.; and Cao, Y

    Sun, Q.; Fang, Y.; Wu, L. Y.; Wang, X.; and Cao, Y. 2023. EVA-CLIP: Improved Training Techniques for CLIP at Scale. ArXiv

    2023

  27. [35]

    Sun, X.; Yang, J.; Sun, M.; and Wang, K. 2016. A benchmark for automatic visual classification of clinical skin disease images. In ECCV

    2016

  28. [36]

    F.; Alabdulmohsin, I.; Parthasarathy, N.; Evans, T.; Beyer, L.; Xia, Y.; Mustafa, B.; et al

    Tschannen, M.; Gritsenko, A.; Wang, X.; Naeem, M. F.; Alabdulmohsin, I.; Parthasarathy, N.; Evans, T.; Beyer, L.; Xia, Y.; Mustafa, B.; et al. 2025. SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features. arXiv pr...

    2025 arXiv

  29. [37]

    S.; and Santosh, K

    Vettoruzzo, A.; Bouguelia, M.; Vanschoren, J.; R \" o gnvaldsson, T. S.; and Santosh, K. 2024. Advances and Challenges in Meta-Learning: A Technical Review. IEEE Trans. Pattern Anal. Mach. Intell. , 46(7): 4763--4779

    2024

  30. [38]

    C.; and Xing, E

    Wang, H.; Ge, S.; Lipton, Z. C.; and Xing, E. P. 2019. Learning Robust Global Representations by Penalizing Local Predictive Power. In NeurIPS

    2019

  31. [39]

    Wei, Z.; Pan, Z.; and Owens, A. 2024. Efficient Vision-Language Pre-training by Cluster Masking. In CVPR

    2024

  32. [40]

    Wu, S.; Zhang, J.; Zeng, P.; Gao, L.; Song, J.; and Shen, H. T. 2025. Skip tuning: Pre-trained vision-language models are effective and efficient adapters themselves. In CVPR

    2025

  33. [41]

    Wu, X.; Zhan, C.; Lai, Y.; Cheng, M.-M.; and Yang, J. 2019. IP102: A Large-Scale Benchmark Dataset for Insect Pest Recognition. In CVPR

    2019

  34. [42]

    A.; Oliva, A.; and Torralba, A

    Xiao, J.; Hays, J.; Ehinger, K. A.; Oliva, A.; and Torralba, A. 2010. SUN database: Large-scale scene recognition from abbey to zoo. In CVPR

    2010

  35. [43]

    Xie, J.; Zhang, Y.; Peng, J.; Huang, Z.; and Cao, L. 2024. TextRefiner: Internal Visual Feature as Efficient Refiner for Vision-Language Models Prompt Tuning. arXiv preprint arXiv:2412.08176

    2024 arXiv

  36. [44]

    E.; Huang, P.; Howes, R.; Sharma, V.; Li, S.; Ghosh, G.; Zettlemoyer, L.; and Feichtenhofer, C

    Xu, H.; Xie, S.; Tan, X. E.; Huang, P.; Howes, R.; Sharma, V.; Li, S.; Ghosh, G.; Zettlemoyer, L.; and Feichtenhofer, C. 2024. Demystifying CLIP Data. In The Twelfth International Conference on Learning Representations, ICLR 2024, Vienna, Austria, May 7-11, 2024

    2024

  37. [45]

    Yang, Y.; Cui, Z.; Xu, J.; Zhong, C.; Zheng, W.-S.; and Wang, R. 2023. Continual learning with Bayesian model based on a fixed pre-trained feature extractor. Visual Intelligence, 1(1): 5

    2023

  38. [46]

    Yao, H.; Zhang, R.; and Xu, C. 2024. TCP: Textual-based Class-aware Prompt tuning for Visual-Language Model. In CVPR

    2024

  39. [47]

    Yu, T.; Lu, Z.; Jin, X.; Chen, Z.; and Wang, X. 2023. Task Residual for Tuning Vision-Language Models. In CVPR

    2023

  40. [48]

    Zhai, X.; Mustafa, B.; Kolesnikov, A.; and Beyer, L. 2023. Sigmoid Loss for Language Image Pre-Training. In ICCV

    2023

  41. [49]

    T.; and Song, J

    Zhang, J.; Wu, S.; Gao, L.; Shen, H. T.; and Song, J. 2024. Dept: Decoupled prompt tuning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    2024

  42. [50]

    Zhang, R.; Zhang, W.; Fang, R.; Gao, P.; Li, K.; Dai, J.; Qiao, Y.; and Li, H. 2022. Tip-Adapter: Training-Free Adaption of CLIP for Few-Shot Classification. In ECCV

    2022

  43. [51]

    C.; and Liu, Z

    Zhou, K.; Yang, J.; Loy, C. C.; and Liu, Z. 2022 a . Conditional Prompt Learning for Vision-Language Models. In CVPR

    2022

  44. [52]

    C.; and Liu, Z

    Zhou, K.; Yang, J.; Loy, C. C.; and Liu, Z. 2022 b . Learning to Prompt for Vision-Language Models. International Journal of Computer Vision, 130(9): 2337--2348

    2022

  45. [53]

    Zhu, B.; Niu, Y.; Han, Y.; Wu, Y.; and Zhang, H. 2023. Prompt-aligned gradient for prompt tuning. In ICCV

    2023

Pith tools

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