Pith. sign in

REVIEW 4 major objections 4 minor 75 references

Simulate, Refocus and Ensemble: An Attention-Refocusing Scheme for Domain Generalization

T0 review · 4 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read This paper claims that refocusing CLIP's attention on simulated target domains raises its accuracy on unseen domains, reporting a 77.9% average on five standard benchmarks.

desk verdict A plausible new CLIP DG pipeline whose headline SOTA claim rests on a 0.6-point average over mixed-protocol baselines with no error bars, so treat Tables I-VI as provisional until seeded comparison. read the letter →

arxiv 2507.12851 v1 pith:I22BWDYI submitted 2025-07-17 cs.CV

Ziyi Wang , Zhi Gao , Jin Chen , Qingjie Zhao , Xinxiao Wu , Jiebo Luo This is my paper
classification cs.CV MSC 68T0768T45
keywords domaingeneralizationCLIPattentionrefocusingvisualprompttuningmulti-sourceopen-setvisiontransformer
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 proposes Simulate, Refocus and Ensemble (SRE), a three-stage scheme to improve CLIP-based domain generalization by steering CLIP's visual attention toward task-relevant, domain-invariant image regions. The first stage creates simulated target domains by applying color jitter, Gaussian blur, and grayscale transformations to source images. The second stage trains an attention-refocuser that adds learned biases to the value matrices of CLIP's self-attention layers, aligning attention maps between source and simulated target images. The third stage ensembles only the refocuser parameters whose attention maps are highly consistent between source and simulated targets, which stabilizes training. On five domain-generalization benchmarks with a ViT-B/16 backbone, SRE reports an average accuracy of 77.9%, including 60.8% on TerraIncognita, surpassing the compared CLIP-based methods.

What carries the argument

The attention-refocuser is the load-bearing component. It contains a token selection module that scores each visual token's task relevance by cosine similarity with a learnable prompt $\sigma$, producing masks $M^s$ and $M^t$, and a decoder $D_\theta$ with one layer per CLIP encoder layer, which converts selected tokens into refocus parameters $e^s_b$ and $e^t_b$. These parameters are added to the input of the value-matrix computation in the $b$-th self-attention layer, $\hat{V}_b = W_{V_b}(X_b + e_b)$, so a second forward pass yields refocused embeddings. The ensemble stage then keeps a weighted average $\theta_a$ of decoder parameters only when the cosine similarity between the masked last-layer attention maps of a source image and its simulated counterpart exceeds an exponentially updated threshold.

What would settle it

Train SRE on the source domains of PACS and evaluate on the same classes rendered with a shift that is not in the augmentation family, such as weather corruptions (snow, frost) or quickdraw-style line sketches; if accuracy falls to or below zero-shot CLIP while a prompt-tuning baseline retains its gain, the simulated target domains are not representative. A second check: compare SRE against its SR+EMA variant on a new benchmark; if equal accuracy appears, the ensemble selection rule is not the active ingredient.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central claim is that CLIP's failure in domain generalization is primarily an attention problem, and that it can be fixed without target data by training a lightweight attention-refocuser to align attention between source images and cheaply simulated target images. The paper reports that this scheme generally outperforms existing CLIP-based domain generalization methods across PACS, VLCS, OfficeHome, TerraIncognita, and DomainNet on three backbones, with the largest margin on the hardest benchmark, TerraIncognita. Single-domain and open-set generalization experiments, along with ablations, are presented as evidence that each stage contributes: refocusing alone improves over zero-shot CLIP, adding simulated domains improves further, and parameter ensembling adds the remaining gain.

Load-bearing premise

The load-bearing assumption is that color jitter, Gaussian blur, and grayscale, applied with unspecified intensity ranges $\phi_i$, produce simulated target domains whose distribution shifts train the refocuser in a way that transfers to real unseen target domains; if that assumption fails, the attention alignment and the ensemble selection criterion are both tuned to the wrong kind of shift.

Editorial extensions

If this is right

  • If SRE's reported accuracy holds, attention refocusing on simulated domains is a viable alternative to prompt tuning for CLIP-based domain generalization, without needing any target-domain data.
  • The large gain on TerraIncognita suggests the method is most valuable when background clutter and environmental shift dominate, since that is where attention drift hurts.
  • Because the refocuser is a small learned module over a frozen CLIP, the scheme can be combined with prompt-based methods or applied to other frozen transformer encoders.
  • The ensemble selection criterion, based on attention consistency between source and simulated target, offers a general training-stabilization recipe for parameter-efficient adapters.

Reading between the lines

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

  • Editorial extension: the same Simulate-Refocus-Ensemble recipe could be transferred to other vision-language models or to video and audio encoders, where the three augmentations would need to be replaced by domain-appropriate shifts such as audio noise or frame-level corruption.
  • Editorial extension: a stronger test of the simulate stage would be to train SRE with augmentation intensities matched to each source domain's own style statistics rather than globally sampled values, since the paper leaves the intensity ranges $\phi_i$ unspecified.
  • Editorial extension: the headline average is carried largely by TerraIncognita, and on VLCS the compared method CLIPCEIL reports higher accuracy, so the claim is best read as specific to high-shift benchmarks rather than universal.
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

4 major / 4 minor

Summary. The manuscript proposes SRE (Simulate, Refocus and Ensemble), a three-stage method for CLIP-based multi-source domain generalization. The Simulate stage applies ColorJitter, GaussianBlur, and GrayScale to source images to create simulated target domains. The Refocus stage trains an attention-refocuser composed of a learnable prompt sigma and a decoder theta that adds learned biases to the value matrices in CLIP's self-attention, using cross-entropy on source and simulated images plus a variance loss on task-relevance masks. The Ensemble stage maintains an exponentially updated threshold and averages decoder weights when source-to-simulated attention consistency exceeds the threshold. Experiments are reported on PACS, VLCS, OfficeHome, TerraIncognita, DomainNet, and NICO++, with additional single-domain and open-set generalization results; the central claim is that SRE generally outperforms state-of-the-art CLIP-based DG methods, with an average of 77.9 on ViT-B/16 across five benchmarks.

Significance. If the reported gains were robust, SRE would provide a practical, relatively lightweight way to improve CLIP's focus on task-relevant regions for unseen domains. The code release, the use of training-domain validation, and the explicit ablations of stages, losses, augmentations, and hyperparameters are strengths. However, the central SOTA claim is not established by the current evidence: Table I compares copied and reproduced baselines under uncontrolled protocols and reports no variances, and the loss ablation has a numerical inconsistency. The method's mechanism description also overstates what is optimized, because the objective contains no attention-map alignment term. These issues are fixable, but they are load-bearing for the paper's main conclusion.

major comments (4)
  1. [Section IV-C, Table I] The claim that SRE generally achieves better results than state-of-the-art methods is not supported as reported. The table mixes numbers copied from original papers (DPL, CSVPT, CLIPCEIL, etc.) with the authors' reproduced baselines (Gallop*, VPT*) and SRE, and no standard deviations or seed counts are given anywhere in the main tables. With different source papers using different validation protocols and training budgets, the 0.6-point average margin over CLIPCEIL (77.9 vs 77.3) is within run-to-run variation for CLIP fine-tuning, especially because SRE is 4.7 points behind CLIPCEIL on VLCS (83.7 vs 88.4). Please rerun all strongest baselines under the identical training-domain-validation protocol and report means and standard deviations over at least three seeds.
  2. [Section III-C, Eqs. (6)-(8)] The paper repeatedly states that the Refocus stage aligns the attention maps of source and simulated target images, but the optimized objective contains no attention-map alignment term. Lcls in Eq. (6) and Lalign in Eq. (7) are cross-entropy losses on the source and simulated images, and Lvar in Eq. (4) only promotes variance within each task-relevance mask; the cosine similarity in Eq. (10) is computed in Section III-D for parameter selection only and is not part of the loss. Thus the described attention-map alignment mechanism is not actually being optimized. Either add an explicit alignment loss on As and At, or revise the description to state that the decoder is trained by classification on simulated domains and the consistency score is only an ensemble-selection criterion.
  3. [Section III-B, Eq. (1)] The augmentation intensity parameters phi_i are said to be uniformly sampled from a predefined range, but the ranges are never specified, and it is unclear whether GrayScale has an intensity parameter at all. Because the simulated target domains are the sole mechanism by which the refocuser sees domain shift, these ranges are a core component of the method and must be reported in Section IV-B for reproducibility.
  4. [Section IV-F-b, Tables XII-XIII] The loss ablation text states that the full model achieves 75.8% average and 59.8% on Terra, whereas Table XIII reports 76.9% average and 60.8% on Terra. In addition, the table numbering is duplicated: two different tables are labeled Table XII and two different tables are labeled Table XIII. Please reconcile the numbers and renumber the tables so the ablation claims can be verified.
minor comments (4)
  1. [Table I and throughout] The backbone names ViT-32/B and ViT-16/B should be written as ViT-B/32 and ViT-B/16, respectively.
  2. [Section IV-G, Figures 8-9] The Grad-CAM visualizations are described qualitatively; please report the quantitative protocol, such as the layer used and whether attention maps are averaged over heads, so the visualizations can be reproduced.
  3. [Figure 7] The caption says the hyperparameter results are averaged over three runs, but no error bars or individual run values are shown; please add them or state the variance in the text.
  4. [Table XV] The memory and latency columns should specify whether memory refers to peak GPU memory and whether latency includes only forward passes; the 0 ms/img entry for ZS-CLIP is misleading because no training is performed.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: SRE's training and ensemble selection use only source and augmented-source data; the SOTA margin issue is a benchmarking protocol concern, not a circular derivation.

full rationale

The Simulate stage (Eq. 1) constructs xt from xs through fixed augmentations; the Refocus stage optimizes Lcls and Lalign (Eqs. 6-8) with source labels, and the Ensemble stage (Eqs. 10-12) selects decoder parameters using attention-map similarity between xs and xt, both quantities available during training. No equation is fitted to test-domain statistics and then reported as a prediction, and no parameter is renamed as a result. The paper's claim of general superiority over CLIPCEIL rests on a small average margin assembled from copied numbers without error bars (Table I), but that is a correctness and evidence issue, not a circular step. Self-citations ([1], [33]) are contextual related-work pointers and carry no load-bearing argument; no uniqueness theorem or ansatz is imported from the authors' prior work. The limitation section acknowledges only complexity and computational cost, which is unrelated to circularity. Hence the derivation chain is self-contained.

Assumptions & free parameters 6 free parameters · 5 assumptions · 1 invented entities

The central claim rests on standard training assumptions and one key modeling choice: that simple photometric augmentations create a good proxy for real domain shift. The unreported augmentation intensity ranges are a notable omission. No free parameter is fitted to test-domain data, so circularity burden is low.

free parameters (6)
  • lambda (variance loss weight) = 0.1
    Selected by validation; sensitivity is shown in Fig. 7 but no automatic tuning is described.
  • omega (ensemble update ratio) = 0.98
    Selected by validation; sensitivity is shown in Fig. 7 but no automatic tuning is described.
  • learning rate for decoder theta = 0.0004
    Set by hand in Section IV-B; no sensitivity analysis is provided.
  • learning rate for prompt sigma = 0.001
    Set by hand in Section IV-B; no sensitivity analysis is provided.
  • augmentation intensity ranges (phi_i) = Not reported
    Section III-B says phi_i is sampled from a predefined range, but the ranges are never given, making the Simulate stage under-specified.
  • training iterations = 2000 for PACS, VLCS, OfficeHome; 5000 for Terra and DomainNet
    Chosen per dataset in Section IV-B; no justification or ablation is provided.
assumptions (5)
  • domain assumption Fine-tuning query or value matrices within attention improves generalization bounds (cited to [51])
    Used in Section III-C to justify modifying value matrices; no derivation is provided in this paper.
  • domain assumption Focusing attention on task-relevant regions improves domain generalization
    Central motivation in Sections I and III; supported only by visualizations and benchmarks, not by a theorem.
  • domain assumption ColorJitter, GaussianBlur, and GrayScale produce simulated target domains representative of real unseen targets
    Section III-B assumes these shifts mimic real domain gaps; validated only empirically on specific benchmarks.
  • domain assumption The last-layer attention map refined by the learned masks (Eq. 9) is a valid indicator of parameter robustness for ensembling
    Section III-D defines the selection criterion on this quantity without theoretical justification.
  • standard math Standard calculus, linear algebra, and probability
    Used throughout the derivation of losses and attention equations.
invented entities (1)
  • Attention-refocuser (learnable prompt sigma and decoder D_theta) independent evidence
    purpose: Adds per-layer biases e_b to CLIP value matrices to redirect attention to task-relevant tokens
    The module is fully specified and evaluated on multiple benchmarks; its behavior is visualized in Figures 8 and 9.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Simulate, Refocus and Ensemble: An Attention-Refocusing Scheme for Domain Generalization." pith.science (2026). https://pith.science/paper/I22BWDYI

@misc{pith2026250712851,
  author       = {Pith},
  title        = {Pith review of: Simulate, Refocus and Ensemble: An Attention-Refocusing Scheme for Domain Generalization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/I22BWDYI}},
  note         = {Machine review of arXiv:2507.12851}
}
read the original abstract

Domain generalization (DG) aims to learn a model from source domains and apply it to unseen target domains with out-of-distribution data. Owing to CLIP's strong ability to encode semantic concepts, it has attracted increasing interest in domain generalization. However, CLIP often struggles to focus on task-relevant regions across domains, i.e., domain-invariant regions, resulting in suboptimal performance on unseen target domains. To address this challenge, we propose an attention-refocusing scheme, called Simulate, Refocus and Ensemble (SRE), which learns to reduce the domain shift by aligning the attention maps in CLIP via attention refocusing. SRE first simulates domain shifts by performing augmentation on the source data to generate simulated target domains. SRE then learns to reduce the domain shifts by refocusing the attention in CLIP between the source and simulated target domains. Finally, SRE utilizes ensemble learning to enhance the ability to capture domain-invariant attention maps between the source data and the simulated target data. Extensive experimental results on several datasets demonstrate that SRE generally achieves better results than state-of-the-art methods. The code is available at: https://github.com/bitPrincy/SRE-DG.

Figures

Figures reproduced from arXiv: 2507.12851 by the authors.

Figure 1
Figure 1. Attention maps of input images generated by CLIP and the proposed [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. An overview of the proposed DG attention-refocusing scheme SRE. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Overview of the three stages of the proposed SRE. The [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: The architecture of the attention-refocuser in the [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: Comparison of the test accuracy between SRE and its variant SR on [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: Results of percentages of available training data on OfficeHome. [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: Results of different weights λ of the variance loss and different update ratios ω for ensembling parameter θa on OfficeHome. To further demonstrate the effectiveness of the proposed Simulate, Refocus and Ensemble scheme, we present visual￾ization results of attention m…
Figure 8
Figure 8. Figure 8: Attention visualization of SRE and several existing methods. [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]
Figure 9
Figure 9. Figure 9: Attention visualization of SRE and several variants. [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

75 extracted references · 68 canonical work pages

  1. [1]

    Meta-causal learning for single domain generalization,

    J. Chen, Z. Gao, X. Wu, and J. Luo, “Meta-causal learning for single domain generalization,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2023, pp. 7683– 7692

    work page 2023

  2. [2]

    Causality inspired representation learning for domain generalization,

    F. Lv, J. Liang, S. Li, B. Zang, C. H. Liu, Z. Wang, and D. Liu, “Causality inspired representation learning for domain generalization,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 8046–8056

    work page 2022

  3. [3]

    Generalizing across domains via cross-gradient training,

    S. Shankar, V . Piratla, S. Chakrabarti, S. Chaudhuri, P. Jyothi, and S. Sarawagi, “Generalizing across domains via cross-gradient training,” in International Conference on Learning Representations (ICLR) , 2018

    work page 2018

  4. [4]

    Domain generalization using causal matching,

    D. Mahajan, S. Tople, and A. Sharma, “Domain generalization using causal matching,” in International Conference on Machine Learning (ICML), 2021, pp. 7313–7324

    work page 2021

  5. [5]

    Learning to generate novel domains for domain generalization,

    K. Zhou, Y . Yang, T. Hospedales, and T. Xiang, “Learning to generate novel domains for domain generalization,” in Proceedings of the Eu- ropean Conference on Computer Vision (ECCV) . Springer, 2020, pp. 561–578

    work page 2020

  6. [6]

    Reducing domain gap by reducing style bias,

    H. Nam, H. Lee, J. Park, W. Yoon, and D. Yoo, “Reducing domain gap by reducing style bias,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2021, pp. 8690– 8699

    work page 2021

  7. [7]

    Learning transferable visual models from natural language supervision,

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

    work page 2021

  8. [8]

    Promptstyler: Prompt-driven style generation for source-free domain generalization,

    J. Cho, G. Nam, S. Kim, H. Yang, and S. Kwak, “Promptstyler: Prompt-driven style generation for source-free domain generalization,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2023, pp. 15 656–15 666

    work page 2023

Show all 75 references
  1. [9]

    Learning common and specific visual prompts for domain generalization,

    A. Li, L. Zhuang, S. Fan, and S. Wang, “Learning common and specific visual prompts for domain generalization,” in Proceedings of the Asian Conference on Computer Vision (ACCV) , 2022, pp. 4260–4275

    2022

  2. [10]

    Domain-unified prompt rep- resentations for source-free domain generalization,

    H. Niu, H. Li, F. Zhao, and B. Li, “Domain-unified prompt rep- resentations for source-free domain generalization,” arXiv preprint arXiv:2209.14926, 2022

    2022 arXiv

  3. [11]

    A closer look at the robustness of contrastive language-image pre-training (clip),

    W. Tu, W. Deng, and T. Gedeon, “A closer look at the robustness of contrastive language-image pre-training (clip),” in Proceedings of the 37th International Conference on Neural Information Processing Systems (NeurIPS), 2024

    2024

  4. [12]

    Test-time adaptation with CLIP reward for zero-shot generalization in vision-language models,

    S. Zhao, X. Wang, L. Zhu, and Y . Yang, “Test-time adaptation with CLIP reward for zero-shot generalization in vision-language models,” in International Conference on Learning Representations (ICLR) , 2024

    2024

  5. [13]

    Achieving plasticity- stability trade-off in continual learning through adaptive orthogonal projection,

    D. Cheng, Y . Hu, N. Wang, D. Zhang, and X. Gao, “Achieving plasticity- stability trade-off in continual learning through adaptive orthogonal projection,” IEEE Transactions on Circuits and Systems for Video Technology (TCSVT), 2025

    2025

  6. [14]

    Progressive negative enhancing contrastive learning for image dehazing and beyond,

    D. Cheng, Y . Li, D. Zhang, N. Wang, J. Sun, and X. Gao, “Progressive negative enhancing contrastive learning for image dehazing and beyond,” IEEE Transactions on Multimedia (TMM) , 2024

    2024

  7. [15]

    Continual all-in-one adverse weather removal with knowledge replay on a unified network structure,

    D. Cheng, Y . Ji, D. Gong, Y . Li, N. Wang, J. Han, and D. Zhang, “Continual all-in-one adverse weather removal with knowledge replay on a unified network structure,” IEEE Transactions on Multimedia (TMM) , 2024

    2024

  8. [16]

    Alpha-clip: A clip model focusing on wherever you want,

    Z. Sun, Y . Fang, T. Wu, P. Zhang, Y . Zang, S. Kong, Y . Xiong, D. Lin, and J. Wang, “Alpha-clip: A clip model focusing on wherever you want,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. 13 019–13 029

    2024

  9. [17]

    Grounded text-to-image synthesis with attention refocusing,

    Q. Phung, S. Ge, and J.-B. Huang, “Grounded text-to-image synthesis with attention refocusing,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2024, pp. 7932– 7942

    2024

  10. [18]

    Toast: Transfer learning via attention steering,

    B. Shi, S. Gai, T. Darrell, and X. Wang, “Toast: Transfer learning via attention steering,” arXiv preprint arXiv:2305.15542, vol. 5, no. 7, p. 13, 2023

    2023 arXiv

  11. [19]

    Primecomposer: Faster progressively combined diffusion for image composition with attention steering,

    Y . Wang, W. Zhang, J. Zheng, and C. Jin, “Primecomposer: Faster progressively combined diffusion for image composition with attention steering,” in Proceedings of the 32nd ACM International Conference on Multimedia (ACM MM) , 2024, pp. 10 824–10 832. 12

    2024

  12. [20]

    xecgnet: Fine-tuning attention map within convolutional neural network to improve detection and explainability of concurrent cardiac arrhythmias,

    J. Yoo, T. J. Jun, and Y .-H. Kim, “xecgnet: Fine-tuning attention map within convolutional neural network to improve detection and explainability of concurrent cardiac arrhythmias,” Computer Methods and Programs in Biomedicine (CMPB) , vol. 208, p. 106281, 2021

    2021

  13. [21]

    Learning to diversify for single domain generalization,

    Z. Wang, Y . Luo, R. Qiu, Z. Huang, and M. Baktashmotlagh, “Learning to diversify for single domain generalization,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) , 2021, pp. 834–843

    2021

  14. [22]

    Learning to learn single domain gen- eralization,

    F. Qiao, L. Zhao, and X. Peng, “Learning to learn single domain gen- eralization,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2020, pp. 12 556–12 565

    2020

  15. [23]

    Representation enhancement-stabilization: Reducing bias-variance of domain general- ization,

    W. Huang, Y . Shi, Z. Xiong, and X. X. Zhu, “Representation enhancement-stabilization: Reducing bias-variance of domain general- ization,” in Proceedings of the European Conference on Computer Vision (ECCV). Springer, 2024, pp. 108–125

    2024

  16. [24]

    Poem: polarization of embeddings for domain- invariant representations,

    S.-Y . Jo and S. W. Yoon, “Poem: polarization of embeddings for domain- invariant representations,” in Proceedings of the AAAI Conference on Artificial Intelligence (AAAI) , vol. 37, no. 7, 2023, pp. 8150–8158

    2023

  17. [25]

    Adversarial invariant feature learning with accuracy constraint for domain generalization,

    K. Akuzawa, Y . Iwasawa, and Y . Matsuo, “Adversarial invariant feature learning with accuracy constraint for domain generalization,” in Euro- pean Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD) . Springer, 2020, pp. 315–331

    2020

  18. [26]

    Grounding stylistic domain generalization with quantitative domain shift measures and synthetic scene images,

    Y . Luo, J. Feinglass, T. Gokhale, K.-C. Lee, C. Baral, and Y . Yang, “Grounding stylistic domain generalization with quantitative domain shift measures and synthetic scene images,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 20...

    2024

  19. [27]

    Self-challenging improves cross-domain generalization,

    Z. Huang, H. Wang, E. P. Xing, and D. Huang, “Self-challenging improves cross-domain generalization,” in Proceedings of the European Conference on Computer Vision (ECCV). Springer, 2020, pp. 124–140

    2020

  20. [28]

    Gradient matching for domain generalization,

    Y . Shi, J. Seely, P. Torr, S. N, A. Hannun, N. Usunier, and G. Syn- naeve, “Gradient matching for domain generalization,” in International Conference on Learning Representations (ICLR) , 2022

    2022

  21. [29]

    Dna: Domain generalization with diversified neural averaging,

    X. Chu, Y . Jin, W. Zhu, Y . Wang, X. Wang, S. Zhang, and H. Mei, “Dna: Domain generalization with diversified neural averaging,” in International Conference on Machine Learning (ICML) . PMLR, 2022, pp. 4010–4034

    2022

  22. [30]

    Sharpness-aware gradient matching for domain generalization,

    P. Wang, Z. Zhang, Z. Lei, and L. Zhang, “Sharpness-aware gradient matching for domain generalization,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023, pp. 3769–3778

    2023

  23. [31]

    Bert: Pre-training of deep bidirectional transformers for language understanding,

    J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova, “Bert: Pre-training of deep bidirectional transformers for language understanding,” in North American Chapter of the Association for Computational Linguistics (ACL), 2019

    2019

  24. [32]

    High- resolution image synthesis with latent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High- resolution image synthesis with latent diffusion models,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion (CVPR), 2022, pp. 10 684–10 695

    2022

  25. [33]

    A hard- to-beat baseline for training-free clip-based adaptation,

    Z. Wang, J. Liang, L. Sheng, R. He, Z. Wang, and T. Tan, “A hard- to-beat baseline for training-free clip-based adaptation,” arXiv preprint arXiv:2402.04087, 2024

    2024 arXiv

  26. [34]

    Data-free generalized zero-shot learning,

    B. Tang, J. Zhang, L. Yan, Q. Yu, L. Sheng, and D. Xu, “Data-free generalized zero-shot learning,” in Proceedings of the AAAI Conference on Artificial Intelligence (AAAI) , vol. 38, no. 6, 2024, pp. 5108–5117

    2024

  27. [35]

    Disentangled prompt representation for domain generalization,

    D. Cheng, Z. Xu, X. Jiang, N. Wang, D. Li, and X. Gao, “Disentangled prompt representation for domain generalization,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. 23 595–23 604

    2024

  28. [36]

    Dual modality prompt tuning for vision-language pre-trained model,

    Y . Xing, Q. Wu, D. Cheng, S. Zhang, G. Liang, P. Wang, and Y . Zhang, “Dual modality prompt tuning for vision-language pre-trained model,” IEEE Transactions on Multimedia (TMM), vol. 26, pp. 2056–2068, 2023

    2023

  29. [37]

    Captions are worth a thousand words: Enhancing product retrieval with pretrained image-to-text models,

    J. Tang, G. McGoldrick, M. Al-Ghossein, and C.-W. Chen, “Captions are worth a thousand words: Enhancing product retrieval with pretrained image-to-text models,” arXiv preprint arXiv:2402.08532 , 2024

    2024 arXiv

  30. [38]

    Embracing language inclusivity and diversity in clip through continual language learning,

    B. Yang, Y . Dai, X. Cheng, Y . Li, A. Raza, and Y . Zou, “Embracing language inclusivity and diversity in clip through continual language learning,” in Proceedings of the AAAI Conference on Artificial Intelli- gence (AAAI), vol. 38, no. 6, 2024, pp. 6458–6466

    2024

  31. [39]

    Clippo: Image-and- language understanding from pixels only,

    M. Tschannen, B. Mustafa, and N. Houlsby, “Clippo: Image-and- language understanding from pixels only,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023, pp. 11 006–11 017

    2023

  32. [40]

    Linearly mapping from image to text space,

    J. Merullo, L. Castricato, C. Eickhoff, and E. Pavlick, “Linearly mapping from image to text space,” in International Conference on Learning Representations (ICLR), 2023

    2023

  33. [41]

    Exploring homogeneous and heterogeneous consistent label associations for unsupervised visible- infrared person reid,

    L. He, D. Cheng, N. Wang, and X. Gao, “Exploring homogeneous and heterogeneous consistent label associations for unsupervised visible- infrared person reid,” International Journal of Computer Vision (IJCV) , pp. 1–20, 2024

    2024

  34. [42]

    Amortized prompt: Lightweight finetuning for clip in domain generalization,

    X. Zhang, Y . Iwasawa, Y . Matsuo, and S. S. Gu, “Amortized prompt: Lightweight finetuning for clip in domain generalization,” arXiv preprint arXiv:2111.12853, vol. 2, no. 3, p. 5, 2021

    2021 arXiv

  35. [43]

    Stylip: Multi-scale style-conditioned prompt learning for clip-based domain generalization,

    S. Bose, A. Jha, E. Fini, M. Singha, E. Ricci, and B. Banerjee, “Stylip: Multi-scale style-conditioned prompt learning for clip-based domain generalization,” in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) , 2024, pp. 5542–5552

    2024

  36. [44]

    Visual prompt tuning,

    M. Jia, L. Tang, B.-C. Chen, C. Cardie, S. Belongie, B. Hariharan, and S.-N. Lim, “Visual prompt tuning,” in Proceedings of the European Conference on Computer Vision (ECCV). Springer, 2022, pp. 709–727

    2022

  37. [45]

    An image is worth 16x16 words: Trans- formers for image recognition at scale,

    A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly, J. Uszkoreit, and N. Houlsby, “An image is worth 16x16 words: Trans- formers for image recognition at scale,” in International Conference on Learni...

    2021

  38. [46]

    Attention is all you need,

    A. Vaswani, N. M. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, and I. Polosukhin, “Attention is all you need,” in Proceedings of the 31st International Conference on Neural Information Processing Systems (NeurIPS) , 2017

    2017

  39. [47]

    Regionclip: Region-based language-image pretraining,

    Y . Zhong, J. Yang, P. Zhang, C. Li, N. Codella, L. H. Li, L. Zhou, X. Dai, L. Yuan, Y . Liet al., “Regionclip: Region-based language-image pretraining,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2022, pp. 16 793–16 803

    2022

  40. [48]

    Masqclip for open-vocabulary universal image segmentation,

    X. Xu, T. Xiong, Z. Ding, and Z. Tu, “Masqclip for open-vocabulary universal image segmentation,” in Proceedings of the IEEE/CVF Inter- national Conference on Computer Vision (ICCV) , 2023, pp. 887–898

    2023

  41. [49]

    Open- vocabulary panoptic segmentation with text-to-image diffusion models,

    J. Xu, S. Liu, A. Vahdat, W. Byeon, X. Wang, and S. De Mello, “Open- vocabulary panoptic segmentation with text-to-image diffusion models,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023, pp. 2955–2966

    2023

  42. [50]

    Tell your model where to attend: Post-hoc attention steering for llms,

    Q. Zhang, C. Singh, L. Liu, X. Liu, B. Yu, J. Gao, and T. Zhao, “Tell your model where to attend: Post-hoc attention steering for llms,” arXiv preprint arXiv:2311.02262, 2023

    2023 arXiv

  43. [51]

    Theoretical insights into fine-tuning attention mechanism: Generalization and optimization,

    X. Yao, H. Qian, X. Hu, G. Xu, and Y . Liu, “Theoretical insights into fine-tuning attention mechanism: Generalization and optimization,” arXiv preprint arXiv:2410.02247 , 2024

    2024 arXiv

  44. [52]

    Attention diversification for domain gen- eralization,

    R. Meng, X. Li, W. Chen, S. Yang, J. Song, X. Wang, L. Zhang, M. Song, D. Xie, and S. Pu, “Attention diversification for domain gen- eralization,” in Proceedings of the European Conference on Computer Vision (ECCV). Springer, 2022, pp. 322–340

    2022

  45. [53]

    Unbiased metric learning: On the utilization of multiple datasets and web images for softening bias,

    C. Fang, Y . Xu, and D. N. Rockmore, “Unbiased metric learning: On the utilization of multiple datasets and web images for softening bias,” in Proceedings of the IEEE International Conference on Computer Vision (ICCV), 2013, pp. 1657–1664

    2013

  46. [54]

    Deeper, broader and artier domain generalization,

    D. Li, Y . Yang, Y .-Z. Song, and T. M. Hospedales, “Deeper, broader and artier domain generalization,” in Proceedings of the IEEE International Conference on Computer Vision (ICCV) , 2017, pp. 5542–5550

    2017

  47. [55]

    Deep hashing network for unsupervised domain adaptation,

    H. Venkateswara, J. Eusebio, S. Chakraborty, and S. Panchanathan, “Deep hashing network for unsupervised domain adaptation,” in Pro- ceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, pp. 5018–5027

    2017

  48. [56]

    Recognition in terra incognita,

    S. Beery, G. Van Horn, and P. Perona, “Recognition in terra incognita,” in Proceedings of the European Conference on Computer Vision (ECCV) , 2018, pp. 456–473

    2018

  49. [57]

    Moment matching for multi-source domain adaptation,

    X. Peng, Q. Bai, X. Xia, Z. Huang, K. Saenko, and B. Wang, “Moment matching for multi-source domain adaptation,” in Proceedings of the IEEE International Conference on Computer Vision (ICCV) , 2019, pp. 1406–1415

    2019

  50. [58]

    The pascal visual object classes (voc) challenge,

    M. Everingham, L. Van Gool, C. K. Williams, J. Winn, and A. Zisser- man, “The pascal visual object classes (voc) challenge,” International Journal of Computer Vision (IJCV) , vol. 88, pp. 303–338, 2010

    2010

  51. [59]

    Labelme: a database and web-based tool for image annotation,

    B. C. Russell, A. Torralba, K. P. Murphy, and W. T. Freeman, “Labelme: a database and web-based tool for image annotation,” International Journal of Computer Vision (IJCV) , vol. 77, pp. 157–173, 2008

    2008

  52. [60]

    Learning generative visual models from few training examples: An incremental bayesian approach tested on 101 object categories,

    L. Fei-Fei, R. Fergus, and P. Perona, “Learning generative visual models from few training examples: An incremental bayesian approach tested on 101 object categories,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshop. IEEE, 2...

    2004

  53. [61]

    Sun database: Large-scale scene recognition from abbey to zoo,

    J. Xiao, J. Hays, K. A. Ehinger, A. Oliva, and A. Torralba, “Sun database: Large-scale scene recognition from abbey to zoo,” in 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2010, pp. 3485–3492. 13

    2010

  54. [62]

    In search of lost domain generalization,

    I. Gulrajani and D. Lopez-Paz, “In search of lost domain generalization,” arXiv preprint arXiv:2007.01434 , 2020

    2007 arXiv

  55. [63]

    Soft prompt generation for domain generalization,

    S. Bai, Y . Zhang, W. Zhou, Z. Luan, and B. Chen, “Soft prompt generation for domain generalization,” in Proceedings of the European Conference on Computer Vision (ECCV). Springer, 2024, pp. 434–450

    2024

  56. [64]

    Maple: Multi-modal prompt learning,

    M. U. Khattak, H. Rasheed, M. Maaz, S. Khan, and F. S. Khan, “Maple: Multi-modal prompt learning,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023, pp. 19 113–19 122

    2023

  57. [65]

    Statistical learning theory,

    V . N. Vapnik, V . Vapnik et al., “Statistical learning theory,” 1998

    1998

  58. [66]

    Domain-adversarial training of neural networks,

    Y . Ganin, E. Ustinova, H. Ajakan, P. Germain, H. Larochelle, F. Lavi- olette, M. March, and V . Lempitsky, “Domain-adversarial training of neural networks,” Journal of Machine Learning Research (JMLR) , vol. 17, no. 59, pp. 1–35, 2016

    2016

  59. [67]

    Deep coral: Correlation alignment for deep domain adaptation,

    B. Sun and K. Saenko, “Deep coral: Correlation alignment for deep domain adaptation,” in Proceedings of the European Conference on Computer Vision (ECCV) . Springer, 2016, pp. 443–450

    2016

  60. [68]

    Domain generalization by mutual- information regularization with pre-trained models,

    J. Cha, K. Lee, S. Park, and S. Chun, “Domain generalization by mutual- information regularization with pre-trained models,” in Proceedings of the European Conference on Computer Vision (ECCV). Springer, 2022, pp. 440–457

    2022

  61. [69]

    Clip-adapter: Better vision-language models with feature adapters,

    P. Gao, S. Geng, R. Zhang, T. Ma, R. Fang, Y . Zhang, H. Li, and Y . Qiao, “Clip-adapter: Better vision-language models with feature adapters,” International Journal of Computer Vision (IJCV) , vol. 132, no. 2, pp. 581–595, 2024

    2024

  62. [70]

    Gallop: Learning global and local prompts for vision-language models,

    M. Lafon, E. Ramzi, C. Rambour, N. Audebert, and N. Thome, “Gallop: Learning global and local prompts for vision-language models,” in Proceedings of the European Conference on Computer Vision (ECCV) . Springer, 2024, pp. 264–282

    2024

  63. [71]

    Clipceil: Domain generalization through clip via channel refinement and image-text alignment,

    X. Yu, S. Yoo, and Y . Lin, “Clipceil: Domain generalization through clip via channel refinement and image-text alignment,” Proceedings of the 37th International Conference on Neural Information Processing Systems (NeurIPS), vol. 37, pp. 4267–4294, 2024

    2024

  64. [72]

    A fourier-based framework for domain generalization,

    Q. Xu, R. Zhang, Y . Zhang, Y . Wang, and Q. Tian, “A fourier-based framework for domain generalization,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021, pp. 14 383–14 392

    2021

  65. [73]

    Robust and generalizable visual representation learning via random convolutions,

    Z. Xu, D. Liu, J. Yang, C. Raffel, and M. Niethammer, “Robust and generalizable visual representation learning via random convolutions,” in International Conference on Learning Representations (ICLR) , 2021

    2021

  66. [74]

    Grad-cam: Visual explanations from deep networks via gradient-based localization,

    R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh, and D. Batra, “Grad-cam: Visual explanations from deep networks via gradient-based localization,” in Proceedings of the IEEE International Conference on Computer Vision (ICCV) , 2017, pp. 618–626

    2017

  67. [75]

    Nico++: Towards better benchmarking for domain generalization,

    X. Zhang, Y . He, R. Xu, H. Yu, Z. Shen, and P. Cui, “Nico++: Towards better benchmarking for domain generalization,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023, pp. 16 036–16 047

    2023

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.