Pith. sign in

REVIEW 3 major objections 3 minor 55 references

TRUST: Leveraging Text Robustness for Unsupervised Domain Adaptation

T0 review · 3 major / 3 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read The paper claims that a vision model can adapt to an unlabeled target domain by learning from target-image captions, with each caption's trustworthiness measured by normalized CLIP similarity.

desk verdict Plausible language-guided UDA method with a genuinely interesting soft-contrastive trick, but the provided full text is corrupted so the SOTA claim rests entirely on the abstract for now. read the letter →

arxiv 2508.06452 v1 pith:WRFFVBGV submitted 2025-08-08 cs.CV cs.LG

Mattia Litrico , Mario Valerio Giuffrida , Sebastiano Battiato , Devis Tuia This is my paper
classification cs.CVcs.LG
keywords unsuperviseddomainadaptationlanguage-guidedpseudo-labelsCLIPsimilarityuncertaintyestimationsoftcontrastivelearningcaptionsupervisionvision-languagealignment
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

TRUST is a method for unsupervised domain adaptation that uses the language modality as the adapting signal. It generates captions for target images, turns them into pseudo-labels, and estimates how much to trust each label from normalized CLIP similarity between the caption and the class text. Wrong labels from low-quality captions are down-weighted, and a caption-guided soft-contrastive loss aligns vision and language without requiring hard positive/negative pairs. The paper reports new state-of-the-art results on the classical DomainNet benchmark and the complex geographical-shift GeoNet benchmark, arguing that text is more stable than pixels under shifts where both backgrounds and object appearances change.

What carries the argument

Three components carry the argument. Caption-conditioned pseudo-labels supply a class signal that pixels cannot reliably give under complex shifts. Normalized CLIP similarity, computed between a caption and class text, acts as an uncertainty estimate that decides how much each pseudo-label is trusted in the classification loss. The soft-contrastive loss compares images through their captions, treating every pair as both positive and negative, so the vision encoder is pulled toward language structure without needing discrete pair selection.

What would settle it

Take a target-domain image, replace the main noun in its caption with a visually similar but semantically wrong noun, and check whether normalized CLIP similarity assigns low confidence and TRUST's accuracy drops accordingly. If accuracy does not drop with caption corruption, the language signal or the uncertainty reweighting is not the active mechanism; conversely, if a shift with systematically misleading captions makes TRUST fall below a vision-only pseudo-label baseline, the trust assumption fails.

Watch

Extended reading notes

Core claim

The central claim is that captions provide a transferable, robust supervision signal for unsupervised domain adaptation. Concretely, the paper proposes to generate a caption for each unlabeled target image, convert that caption into a pseudo-label, and estimate per-sample uncertainty as the normalized CLIP similarity between the caption and the class label text. The classification loss is reweighted by this uncertainty so that pseudo-labels from low-quality captions contribute less. On top of that, TRUST adds a multimodal soft-contrastive loss in which every pair of target images serves simultaneously as a positive and a negative pair, with attraction and repulsion strength proportional to t

Load-bearing premise

The load-bearing premise is that target-image captions are accurate enough, and normalized CLIP similarity is calibrated enough, for language-derived signals to identify correct classes and reliably down-weight wrong pseudo-labels.

Editorial extensions

If this is right

  • If TRUST is correct, a vision encoder can be adapted to a target domain without any target labels, using only captions produced for target images.
  • The uncertainty reweighting should make the method tolerant of partially wrong captions, reducing the need for manual caption filtering.
  • The caption-guided soft-contrastive loss removes the need to select discrete positive and negative pairs, a known failure point in unsupervised domain adaptation.
  • On the benchmarks tested, TRUST is claimed to outperform prior UDA approaches, including under geographical shift where pixel-only methods struggle.

Reading between the lines

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

  • The caption-similarity-as-strength contrastive principle is not tied to classification: it could carry over to retrieval, detection, or segmentation with weak text descriptions, though the paper only demonstrates classification accuracy.
  • Because the uncertainty signal comes from CLIP's image-text alignment, TRUST likely inherits CLIP's blind spots; on a domain where CLIP is confidently wrong, the reweighting could amplify errors rather than suppress them.
  • A natural stress test is to corrupt target captions at a controlled rate and observe accuracy; the paper's mechanism predicts monotone degradation with caption quality, while a vision-only baseline would be unaffected by caption corruption.
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

3 major / 3 minor

Summary. The paper proposes TRUST, an unsupervised domain adaptation method that uses captions of target images to generate pseudo-labels, estimates pseudo-label uncertainty from normalized CLIP similarity scores, and reweights the classification loss accordingly. It also introduces a multimodal soft-contrastive loss in which every pair of images is both positive and negative, with attraction/repulsion strengths proportional to caption similarity, avoiding hard pair selection. The reported claim is state-of-the-art accuracy on DomainNet and GeoNet. The submitted full text is largely unreadable due to encoding corruption, so the abstract and a few equation fragments are the only reliably assessable content.

Significance. If the empirical claims hold, the paper offers a plausible and conceptually interesting mechanism for using language robustness in UDA: caption-derived pseudo-labels, uncertainty-weighted by CLIP similarity, plus a soft-contrastive loss that avoids brittle pair selection. The soft-contrastive formulation is elegant and could be reused beyond this specific setting. However, the current submission does not make the evidence available for verification: there are no visible experimental tables, ablations, variance estimates, or protocol descriptions. The central SOTA claim is therefore unsupported in the manuscript as received. The method also rests on strong assumptions about CLIP similarity calibration under target shift that are not analyzed.

major comments (3)
  1. [Abstract / Experimental evaluation] The central claim—'outperforms previous methods, setting the new state-of-the-art on classical (DomainNet) and complex (GeoNet) domain shifts'—is empirical and load-bearing. The provided manuscript contains no readable results section, no numeric comparison table, no standard deviations or seeds, no ablations of the uncertainty reweighting or the soft-contrastive loss, and no dataset/protocol description. As a result, the SOTA claim is unsupported by any verifiable evidence. A complete experimental section with comparisons, ablations, and variance is required.
  2. [Uncertainty estimation (abstract; reweighted classification loss)] The method reweights pseudo-label classification losses by normalized CLIP similarity scores. For the reweighting to suppress wrong pseudo-labels, these scores must be monotonically related to the probability that the pseudo-label is correct in the target domain. The manuscript does not provide any calibration analysis (e.g., reliability diagrams, accuracy-vs-confidence plots for pseudo-labels) and does not discuss how calibration behaves under large geographic or complex shifts. Without this, the reported gains could come from the soft-contrastive loss or the CLIP backbone rather than from the uncertainty reweighting. Add an explicit calibration study or an ablation that replaces the reweighting with uniform weights.
  3. [Method / initialization dependency] The caption-based pseudo-labels and the CLIP-similarity uncertainty scores both originate from the same frozen CLIP model, and the vision encoder is typically CLIP-initialized. This makes the adaptation signal dependent on CLIP's image-text alignment, so the claim that 'language robustness' causes the improvement is potentially confounded. Concretely, the method could inherit CLIP's biases and limitations rather than add robustness. To support the causal claim, the authors should provide an ablation with a non-CLIP vision backbone or with pseudo-labels produced by an independent captioning model, and should report performance when the uncertainty weighting is removed.
minor comments (3)
  1. [Full text readability] The submitted full text is heavily corrupted (mojibake/encoding artifacts), with only fragments of equations and repeated figure captions legible. This makes the technical description and the experiments impossible to verify. A clean PDF or LaTeX source is needed for review.
  2. [Equations and notation] From the legible fragments, the loss notation (L_CE, L_SCL, etc.) is not fully introduced in readable form. Please ensure every symbol in the final loss expressions is defined and that the combination weights are explicitly stated.
  3. [Reproducibility] The paper says code will be available upon acceptance. For a method with several interacting components (pseudo-label generation, uncertainty reweighting, soft-contrastive loss), a reproducibility statement with hyperparameters, architectures, and training schedules would substantially increase confidence.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: TRUST relies on CLIP for pseudo-labels and uncertainty, but the final predictions are evaluated against external target labels and benchmarks.

full rationale

The paper's derivation chain is: (1) generate caption-based pseudo-labels for target images; (2) estimate their uncertainty via normalized CLIP similarity; (3) reweight the classification loss by that uncertainty; and (4) add a caption-similarity-weighted soft-contrastive loss. The only potentially circular-looking point is that pseudo-labels and uncertainty both derive from CLIP similarity scores, making the uncertainty a confidence measure of the same model that produced the pseudo-labels. But this is a standard self-training/confidence-weighting design, not a logical reduction: the pseudo-labels are predictions about external target classes, and accuracy is measured against those external labels. CLIP is a fixed, externally grounded model, not a parameter fitted to the target data and not a conclusion imported from the paper's own prior work. No equation in the readable text defines the target prediction as identical to an input by construction, no fitted parameter is renamed as a prediction, and no load-bearing self-citation or imported uniqueness theorem is present. The state-of-the-art claim is an empirical comparison on DomainNet and GeoNet, which is externally falsifiable. Therefore no significant circularity is found.

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

The ledger reflects that only the abstract is readable. TRUST leans on CLIP as an external foundation model; the key unverified inputs are caption reliability and CLIP-similarity calibration. Hyperparameters are not visible. No new physical or ontological entities are introduced; the named components are algorithmic constructs.

free parameters (2)
  • Uncertainty weighting coefficients and reweighting functional form = not reported
    The abstract says uncertainty 'reweights the classification loss' but gives no functional form, temperature, or scaling constants. Corpus unreadable, so values cannot be extracted.
  • Soft-contrastive loss mixing strength = not reported
    The relative weight of the contrastive loss against the classification loss, and any caption-similarity temperature, are not visible in the abstract and cannot be read in the body.
assumptions (4)
  • domain assumption Text/caption embeddings are substantially more robust to domain shift than visual embeddings, so captions yield reliable pseudo-labels in the target domain.
    This is the core motivation stated in the abstract ('language modality can help... exhibiting more robustness'); if captions are unreliable in the target domain, pseudo-labels propagate errors.
  • domain assumption Normalized CLIP similarity scores are a valid, approximately calibrated measure of pseudo-label uncertainty.
    The uncertainty strategy assumes caption-image similarity reflects pseudo-label correctness; the abstract gives no evidence of calibration under domain shift.
  • domain assumption Every target image has an available caption from which a pseudo-label can be generated.
    Pseudo-labels are 'generated from their captions'; the abstract does not state whether captions come from existing metadata or from a captioning system, or how caption failures are handled.
  • domain assumption DomainNet and GeoNet are used under standard public benchmark protocols.
    The SOTA claim presumes standard splits and label spaces; unverifiable from the abstract alone.

how reviews work

0 comments
Cite this review

Pith. "Pith review of TRUST: Leveraging Text Robustness for Unsupervised Domain Adaptation." pith.science (2026). https://pith.science/paper/WRFFVBGV

@misc{pith2026250806452,
  author       = {Pith},
  title        = {Pith review of: TRUST: Leveraging Text Robustness for Unsupervised Domain Adaptation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WRFFVBGV}},
  note         = {Machine review of arXiv:2508.06452}
}
read the original abstract

Recent unsupervised domain adaptation (UDA) methods have shown great success in addressing classical domain shifts (e.g., synthetic-to-real), but they still suffer under complex shifts (e.g. geographical shift), where both the background and object appearances differ significantly across domains. Prior works showed that the language modality can help in the adaptation process, exhibiting more robustness to such complex shifts. In this paper, we introduce TRUST, a novel UDA approach that exploits the robustness of the language modality to guide the adaptation of a vision model. TRUST generates pseudo-labels for target samples from their captions and introduces a novel uncertainty estimation strategy that uses normalised CLIP similarity scores to estimate the uncertainty of the generated pseudo-labels. Such estimated uncertainty is then used to reweight the classification loss, mitigating the adverse effects of wrong pseudo-labels obtained from low-quality captions. To further increase the robustness of the vision model, we propose a multimodal soft-contrastive learning loss that aligns the vision and language feature spaces, by leveraging captions to guide the contrastive training of the vision model on target images. In our contrastive loss, each pair of images acts as both a positive and a negative pair and their feature representations are attracted and repulsed with a strength proportional to the similarity of their captions. This solution avoids the need for hardly determining positive and negative pairs, which is critical in the UDA setting. Our approach outperforms previous methods, setting the new state-of-the-art on classical (DomainNet) and complex (GeoNet) domain shifts. The code will be available upon acceptance.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

55 extracted references · 46 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]

    Caron, M.; Bojanowski, P.; Joulin, A.; and Douze, M. 2018. Deep Clustering for Unsupervised Learning of Visual Features. In European Conference on Computer Vision

    work page 2018

  4. [4]

    Chen, D.; Wang, D.; Darrell, T.; and Ebrahimi, S. 2022 a . Contrastive Test-time Adaptation. In CVPR

    work page 2022

  5. [5]

    Chen, L.; Chen, H.; Wei, Z.; Jin, X.; Tan, X.; Jin, Y.; and Chen, E. 2022 b . Reusing the Task-specific Classifier as a Discriminator: Discriminator-free Adversarial Domain Adaptation. 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 7171--7180

    work page 2022

  6. [6]

    Chen, S.; Zhang, Y.; Jiang, W.; Lu, J.; and Zhang, Y. 2024. VLLaVO: Mitigating Visual Gap through LLMs. arXiv:2401.03253

    work page Pith review arXiv 2024

  7. [7]

    Chen, T.; Kornblith, S.; Norouzi, M.; and Hinton, G. 2020 a . A simple framework for contrastive learning of visual representations. In Proceedings of the 37th International Conference on Machine Learning, ICML'20. JMLR.org

    work page 2020

  8. [8]

    Chen, T.; Kornblith, S.; Swersky, K.; Norouzi, M.; and Hinton, G. 2020 b . Big self-supervised models are strong semi-supervised learners. In Proceedings of the 34th International Conference on Neural Information Processing Systems, NIPS '20. Red Hook, NY, USA: Curran Associates Inc. ISBN 9781713829546

    work page 2020

Show all 55 references
  1. [9]

    Chen, X.; and He, K. 2020. Exploring Simple Siamese Representation Learning. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 15745--15753

    2020

  2. [10]

    Cho, J.; Nam, G.; Kim, S.; Yang, H.; and Kwak, S. 2023. PromptStyler: Prompt-driven Style Generation for Source-free Domain Generalization. In 2023 IEEE/CVF International Conference on Computer Vision (ICCV), 15656--15666. Los Alamitos, CA, USA: IEEE Computer Society

    2023

  3. [11]

    Chuang, C.-Y.; Robinson, J.; Lin, Y.-C.; Torralba, A.; and Jegelka, S. 2020. Debiased Contrastive Learning. ArXiv, abs/2007.00224

    2020 arXiv

  4. [12]

    Devlin, J.; Chang, M.-W.; Lee, K.; and Toutanova, K. 2019. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. In North American Chapter of the Association for Computational Linguistics

    2019

  5. [13]

    Du, Z.; Li, J.; Su, H.; Zhu, L.; and Lu, K. 2021. Cross-Domain Gradient Discrepancy Minimization for Unsupervised Domain Adaptation. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 3936--3945

    2021

  6. [14]

    E.; Raghunathan, A.; and Rohrbach, A

    Dunlap, L.; Mohri, C.; Guillory, D.; Zhang, H.; Darrell, T.; Gonzalez, J. E.; Raghunathan, A.; and Rohrbach, A. 2023. Using Language to Extend to Unseen Domains. arXiv:2210.09520

    2023 arXiv

  7. [15]

    Ganin, Y.; and Lempitsky, V. 2015. Unsupervised Domain Adaptation by Backpropagation. In Proceedings of the 32nd International Conference on International Conference on Machine Learning - Volume 37, ICML'15, 1180–1189. JMLR.org

    2015

  8. [16]

    Ge, C.; Huang, R.; Xie, M.; Lai, Z.; Song, S.; Li, S.; and Huang, G. 2022. Domain Adaptation via Prompt Learning. IEEE transactions on neural networks and learning systems, PP

    2022

  9. [17]

    Goyal, S.; Kumar, A.; Garg, S.; Kolter, Z.; and Raghunathan, A. 2023. Finetune like you pretrain: Improved finetuning of zero-shot vision models. In 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 19338--19347. Los Alamitos, CA, USA: IEEE Computer Society

    2023

  10. [18]

    He, K.; Fan, H.; Wu, Y.; Xie, S.; and Girshick, R. B. 2019. Momentum Contrast for Unsupervised Visual Representation Learning. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 9726--9735

    2019

  11. [19]

    Hoffman, J.; Tzeng, E.; Park, T.; Zhu, J.-Y.; Isola, P.; Saenko, K.; Efros, A.; and Darrell, T. 2018. C y CADA : Cycle-Consistent Adversarial Domain Adaptation. In Dy, J.; and Krause, A., eds., Proceedings of the 35th International Conference on Machine Learning, volume 80 of ...

    2018

  12. [20]

    Huang, Z.; Zhou, A.; Lin, Z.; Cai, M.; Wang, H.; and Lee, Y. J. 2023. A Sentence Speaks a Thousand Images: Domain Generalization through Distilling CLIP with Language Guidance. 2023 IEEE/CVF International Conference on Computer Vision (ICCV), 11651--11661

    2023

  13. [21]

    V.; Sung, Y.-H.; Li, Z.; and Duerig, T

    Jia, C.; Yang, Y.; Xia, Y.; Chen, Y.-T.; Parekh, Z.; Pham, H.; Le, Q. V.; Sung, Y.-H.; Li, Z.; and Duerig, T. 2021. Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision. In International Conference on Machine Learning

    2021

  14. [22]

    B.; Pion, N.; Weinzaepfel, P.; and Larlus, D

    Kalantidis, Y.; Sariyildiz, M. B.; Pion, N.; Weinzaepfel, P.; and Larlus, D. 2020. Hard Negative Mixing for Contrastive Learning. ArXiv, abs/2010.01028

    2020 arXiv

  15. [23]

    Kalluri, T.; and Chandraker, M. 2022. Cluster-to-adapt: Few Shot Domain Adaptation for Semantic Segmentation across Disjoint Labels. In 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 4120--4130. Los Alamitos, CA, USA: IEEE Computer Society

    2022

  16. [24]

    Kalluri, T.; Majumder, B.; and Chandraker, M. 2024. Tell, Don`t Show! Language Guidance Eases Transfer Across Domains in Images and Videos. ICML

    2024

  17. [25]

    Kalluri, T.; Sharma, A.; and Chandraker, M. 2022. MemSAC: Memory Augmented Sample Consistency for Large Scale Domain Adaptation. ArXiv, abs/2207.12389

    2022 arXiv

  18. [26]

    Kalluri, T.; Xu, W.; and Chandraker, M. 2023. GeoNet: Benchmarking Unsupervised Adaptation across Geographies. CVPR

    2023

  19. [27]

    Kang, G.; Jiang, L.; Wei, Y.; Yang, Y.; and Hauptmann, A. 2022. Contrastive Adaptation Network for Single- and Multi-Source Domain Adaptation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(4): 1793--1804

    2022

  20. [28]

    Khosla, P.; Teterwak, P.; Wang, C.; Sarna, A.; Tian, Y.; Isola, P.; Maschinot, A.; Liu, C.; and Krishnan, D. 2020. Supervised contrastive learning. In Proceedings of the 34th International Conference on Neural Information Processing Systems, NIPS '20. Red Hook, NY, USA: Curran...

    2020

  21. [29]

    P.; Li, X.; Fu, K

    Lai, Z.; Vesdapunt, N.; Zhou, N.; Wu, J.; Huynh, C. P.; Li, X.; Fu, K. K.; and Chuah, C.-N. 2023. PADCLIP: Pseudo-labeling with Adaptive Debiasing in CLIP for Unsupervised Domain Adaptation. In 2023 IEEE/CVF International Conference on Computer Vision (ICCV), 16109--16119

    2023

  22. [30]

    Li, J.; Li, D.; Savarese, S.; and Hoi, S. 2023. BLIP-2: bootstrapping language-image pre-training with frozen image encoders and large language models. In Proceedings of the 40th International Conference on Machine Learning, ICML'23. JMLR.org

    2023

  23. [31]

    Li, J.; Socher, R.; and Hoi, S. C. H. 2020. DivideMix: Learning with Noisy Labels as Semi-supervised Learning. ArXiv, abs/2002.07394

    2020 arXiv

  24. [32]

    H.; Liang, J.; Qin, C.; and Li, W

    Li, S.; Xie, M.; Lv, F.; Liu, C. H.; Liang, J.; Qin, C.; and Li, W. 2021. Semantic Concentration for Domain Adaptation. 2021 IEEE/CVF International Conference on Computer Vision (ICCV), 9082--9091

    2021

  25. [33]

    Li, X.; Li, Y.; Du, Z.; Li, F.; Lu, K.; and Li, J. 2024. Split to Merge: Unifying Separated Modalities for Unsupervised Domain Adaptation. In 2024 IEEE Conference on Computer Vision and Pattern Recognition

    2024

  26. [34]

    Litrico, M.; Del Bue, A.; and Morerio, P. 2023. Guiding Pseudo-labels with Uncertainty Estimation for Source-free Unsupervised Domain Adaptation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

    2023

  27. [35]

    Liu, G.; and Wang, Y. 2023. TDG: Text-guided Domain Generalization. arXiv:2308.09931

    2023 arXiv

  28. [36]

    Long, M.; Cao, Z.; Wang, J.; and Jordan, M. I. 2018. Conditional adversarial domain adaptation. In Proceedings of the 32nd International Conference on Neural Information Processing Systems, NIPS'18, 1647–1657. Red Hook, NY, USA: Curran Associates Inc

    2018

  29. [37]

    Long, M.; Zhu, H.; Wang, J.; and Jordan, M. I. 2017. Deep transfer learning with joint adaptation networks. In Proceedings of the 34th International Conference on Machine Learning - Volume 70, ICML'17, 2208–2217. JMLR.org

    2017

  30. [38]

    Mahajan, D.; Girshick, R.; Ramanathan, V.; He, K.; Paluri, M.; Li, Y.; Bharambe, A.; and van der Maaten, L. 2018. Exploring the Limits of Weakly Supervised Pretraining. In Computer Vision – ECCV 2018: 15th European Conference, Munich, Germany, September 8-14, 2018, Proceedings...

    2018

  31. [39]

    Peng, X.; Bai, Q.; Xia, X.; Huang, Z.; Saenko, K.; and Wang, B. 2019. Moment matching for multi-source domain adaptation. In Proceedings of the IEEE International Conference on Computer Vision, 1406--1415

    2019

  32. [40]

    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 Meila, M.; and Zhang, T., eds., Proceedings of...

    2021

  33. [41]

    Saito, K.; Watanabe, K.; Ushiku, Y.; and Harada, T. 2017. Maximum Classifier Discrepancy for Unsupervised Domain Adaptation. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 3723--3732

    2017

  34. [42]

    Sanh, V.; Debut, L.; Chaumond, J.; and Wolf, T. 2019. DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter. ArXiv, abs/1910.01108

    2019 arXiv

  35. [43]

    Sharma, A.; Kalluri, T.; and Chandraker, M. 2021. Instance Level Affinity-Based Transfer for Unsupervised Domain Adaptation. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 5357--5367

    2021

  36. [44]

    Sun, T.; Lu, C.; Zhang, T.; and Ling, H. 2022. Safe Self-Refinement for Transformer-based Domain Adaptation. In 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 7181--7190. Los Alamitos, CA, USA: IEEE Computer Society

    2022

  37. [45]

    Tan, S.; Peng, X.; and Saenko, K. 2020. Class-Imbalanced Domain Adaptation: An Empirical Odyssey. In Computer Vision – ECCV 2020 Workshops: Glasgow, UK, August 23–28, 2020, Proceedings, Part I, 585–602. Berlin, Heidelberg: Springer-Verlag. ISBN 978-3-030-66414-5

    2020

  38. [46]

    Tang, S.; Su, W.; Ye, M.; and Zhu, X. 2023. Source-Free Domain Adaptation with Frozen Multimodal Foundation Model. 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 23711--23720

    2023

  39. [47]

    van den Oord, A.; Li, Y.; and Vinyals, O. 2018. Representation Learning with Contrastive Predictive Coding. ArXiv, abs/1807.03748

    2018 arXiv

  40. [48]

    Wang, Z.; Zhang, L.; Wang, L.; and Zhu, M. 2024. LanDA: Language-Guided Multi-Source Domain Adaptation. ArXiv, abs/2401.14148

    2024 arXiv

  41. [49]

    Wei, G.; Lan, C.; Zeng, W.; Zhang, Z.; and Chen, Z. 2024. ToAlign: task-oriented alignment for unsupervised domain adaptation. In Proceedings of the 35th International Conference on Neural Information Processing Systems, NIPS '21. Red Hook, NY, USA: Curran Associates Inc. ISBN...

    2024

  42. [50]

    Wu, C.; Wu, F.; and Huang, Y. 2021. Rethinking InfoNCE: How Many Negative Samples Do You Need? ArXiv, abs/2105.13003

    2021 arXiv

  43. [51]

    Xu, R.; Li, G.; Yang, J.; and Lin, L. 2019. Larger Norm More Transferable: An Adaptive Feature Norm Approach for Unsupervised Domain Adaptation. In 2019 IEEE/CVF International Conference on Computer Vision (ICCV), 1426--1435. Los Alamitos, CA, USA: IEEE Computer Society

    2019

  44. [52]

    Xu, T.; Chen, W.; Wang, P.; Wang, F.; Li, H.; and Jin, R. 2021. CDTrans: Cross-domain Transformer for Unsupervised Domain Adaptation. ArXiv, abs/2109.06165

    2021 arXiv

  45. [53]

    Zhang, Y.; Liu, T.; Long, M.; and Jordan, M. I. 2019. Bridging Theory and Algorithm for Domain Adaptation. In International Conference on Machine Learning

    2019

  46. [54]

    Zhekai Du, F. L. K. L. L. Z. J. L., Xinyao Li. 2024. Domain-Agnostic Mutual Prompting for Unsupervised Domain Adaptation. In IEEE Conference on Computer Vision and Pattern Recognition

    2024

  47. [55]

    Zhu, J.; Bai, H.; and Wang, L. 2023. Patch-Mix Transformer for Unsupervised Domain Adaptation: A Game Perspective. arXiv:2303.13434

    2023 arXiv

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.