Pith. sign in

REVIEW 4 major objections 6 minor 28 references

DCFS: Continual Test-Time Adaptation via Dual Consistency of Feature and Sample

T0 review · 4 major / 6 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 A feature-disentangling continual test-time adaptation method matches or beats the best prior results on three standard corruption benchmarks.

desk verdict A plausible, well-engineered CTTA method with competitive benchmark numbers, but the core disentanglement claim is unvalidated and the gains over PALM are within noise without seeds or error bars. read the letter →

arxiv 2508.20516 v1 pith:FEYJUA6A submitted 2025-08-28 cs.CV

Wenting Yin , Han Sun , Xinru Meng , Ningzhong Liu , Huiyu Zhou This is my paper · ORCID
classification cs.CV
keywords Continualtest-timeadaptationFeaturedisentanglementPseudo-labelnoiseConfidence-awarelearningCorruptionrobustnessCoordinateattentionMixupErroraccumulation
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 DCFS, a method for continual test-time adaptation (CTTA) that keeps adapting a pretrained model to an online stream of corrupted target images without access to source data. The central proposal is to split each target image's feature map into a semantic-related part and a domain-related part using a coordinate-attention mask, train a frozen domain classifier and an updating semantic classifier on those parts, and enforce consistency between sub-feature predictions and whole-feature predictions. On top of that, the paper weights self-supervised learning per sample by a confidence score estimated from a truncated Gaussian model of batch predictions, with uniform alignment to prevent class imbalance. The method reports mean error rates of 15.5% on CIFAR10-C, 30.0% on CIFAR100-C, and 59.8% on ImageNet-C, matching or beating the previous best continual test-time adaptation results on all three benchmarks. The claim is that this reduces pseudo-label noise and error accumulation, the main failure mode of online self-training.

What carries the argument

The load-bearing construction is the coordinate-attention mask $A(F)$ in Eq. (1), which factorizes the feature $F = g(x)$ into $F_S = F \cdot A(F)$ and $F_D = F \cdot (1-A(F))$. The paper treats $F_S$ as semantic-related and $F_D$ as domain-related, feeds them to two classifiers $h_S$ (updated during adaptation) and $h_D$ (frozen source classifier), and uses their summed predictions as the disentangled pseudo-label. The matching losses are single-sample and Mixup consistency between that pseudo-label and the whole-feature output, plus a classifier-discrepancy regularizer $\|W_1^\top W_2\|_1$ that pushes the two classifiers apart. The second mechanism is the truncated-Gaussian confidence weight $\lambda(y)$ in Eq. (11), with batch statistics carried by exponential moving averages and uniform alignment to rebalance categories, which decides how much each sample's self-supervised prediction contributes.

What would settle it

A controlled experiment that replaces $A(F)$ with a random-but-similar-capacity mask, or with the identity (no split), while keeping all losses and the total parameter count fixed, would reveal whether the reported gains come from the disentanglement itself; if the random split matches DCFS's error rates, the semantic/domain decomposition is not doing the work. A second check is to visualize $F_S$ and $F_D$ across corruption types: the semantic part should stay stable across corruptions and the domain part should track the corruption identity.

Watch

Extended reading notes

Core claim

DCFS's central claim is that pseudo-label noise and error accumulation in continual test-time adaptation can be reduced by construction: instead of trusting the whole-feature prediction, the model uses a frozen source classifier to absorb domain-specific features while an updated classifier specializes on semantic features, so the two predictions disagree mainly on domain variance. The consistency losses, both single-sample and Mixup-based, force the whole-feature predictor to agree with the disentangled sub-feature predictions, and a confidence-weighted sample loss down-weights unreliable pseudo-labels. The reported consequence is that on the CIFAR10-C, CIFAR100-C, and ImageNet-C continual benchmarks, DCFS achieves state-of-the-art or tied-best mean error rates of 15.5%, 30.0%, and 59.8% respectively at corruption severity level 5.

Load-bearing premise

The load-bearing premise is that the coordinate-attention mask $A(F)$ in Eq. (1) genuinely separates each feature map into a semantic part and a domain part, and that keeping the domain classifier frozen while updating the semantic classifier makes the two classifiers specialize in those two parts; the paper gives no direct check of this decomposition.

Editorial extensions

If this is right

  • On CIFAR10-C, DCFS reaches a mean error of 15.5%, tying PALM and improving on CoTTA by 0.7% at severity level 5.
  • On CIFAR100-C, DCFS reaches a mean error of 30.0%, the best reported result across the 15 corruption types.
  • On ImageNet-C, DCFS reaches a mean error of 59.8%, the best listed result, ahead of PALM by 0.3%.
  • Ablation on CIFAR10-C shows the three components, feature disentanglement consistency, classifier discrepancy maximization, and sample consistency learning, together reduce mean error from 43.5% to 15.5%, and performance is insensitive to the two balancing hyperparameters.
  • If the claim holds, the consistency and confidence-weighting mechanisms should slow the accumulation of errors over long corruption streams, which is the core practical requirement for continual test-time adaptation.

Reading between the lines

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

  • The feature-disentanglement claim hinges on $A(F)$ actually separating semantic and domain cues; if a random mask of similar structure produced the same gains, the reported improvement would be attributable to added capacity and regularization rather than to disentanglement, and the paper does not include such a control.
  • A natural test beyond the paper is to apply DCFS to gradual or mixed-severity shift sequences, where the confidence statistics in Eq. (12)-(13) would need to track a shifting distribution rather than a fixed corruption level.
  • The truncated-Gaussian confidence weighting plus uniform alignment is a general recipe for pseudo-label selection, so it could be evaluated standalone against fixed-threshold filtering in any online self-training pipeline.
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 / 6 minor

Summary. The paper proposes DCFS, a continual test-time adaptation method that combines feature disentanglement, dual-classifier consistency, and confidence-weighted self-supervised learning. A coordinate-attention mask splits the feature map into semantic-related and domain-related parts; a frozen source classifier and an updated semantic classifier produce two predictions whose sum is aligned with the whole-feature prediction under a cross-entropy loss. A classifier-discrepancy term encourages the two classifiers to use different parameters, and mixup is applied on the disentangled predictions. Sample-level confidence is estimated with an EMA-updated truncated Gaussian and used to weight a self-supervised cross-entropy loss on augmented views, with uniform alignment to balance pseudo-label classes. Experiments on CIFAR10-C, CIFAR100-C, and ImageNet-C under the continual protocol report mean error rates of 15.5%, 30.0%, and 59.8%, respectively, placing DCFS at or near the top of the reported comparison.

Significance. Strengths: the paper evaluates on three standard CTTA benchmarks, includes an ablation of the loss components and a sensitivity study of the two main trade-off parameters, and compares with a recent set of baselines including PALM. The method is simple enough to reimplement from the text. However, the central scientific claim—that the mask in Eq. (1) genuinely separates semantic from domain information—is not validated, and the empirical advantages over PALM are 0.0–0.3 percentage points, which are not supported by any repeated-run statistics. If the mechanism were validated and the results confirmed over multiple seeds, DCFS would be a useful addition to the CTTA literature, but the current evidence does not establish that the reported gains arise from the proposed disentanglement rather than from added capacity and regularization.

major comments (4)
  1. [Section 3.1, Eq. (1)] The semantic/domain decomposition is the load-bearing assumption of the paper, but it is never validated. The coordinate-attention mask A(F) is an unconstrained differentiable output; a degenerate mask (saturating to 0 or 1) or even a random fixed mask would define the same loss family, so the reported gains could come from the extra classifier and regularization rather than from true disentanglement. Please provide at least one of: (a) statistics on the mask values (e.g., fraction of near-0/1 entries) over a target domain stream, (b) a control experiment with a random or fixed mask, (c) qualitative evidence such as attention maps showing that FS localizes semantics while FD captures style/domain cues. Table 4 ablates whole losses and does not address this. In addition, the paper calls the frozen source classifier h_D a 'domain classifier' and says it 'preserves domain-related information,' but the source classifier was trained on semantic labels, so this role is not grounded.
  2. [Section 3.1, Eqs. (3) and (6)-(7)] The target P = PS + PD (and the mixed-label ey = ρ(P_iS+P_iD)+(1−ρ)(P_jS+P_jD)) is used inside a cross-entropy loss without renormalization. Since PS and PD are softmax outputs, the sum has total mass 2 (and ey also has mass 2), which changes the scale and gradient of the CE term and contradicts the phrase 'averaged' used in the text. Please specify the exact normalization and, if normalization was applied in the experiments, state it; if not, re-run with normalized targets and check whether the results change.
  3. [Tables 1-3] The reported advantages over the closest baseline PALM are 0.0 points on CIFAR10-C, 0.1 points on CIFAR100-C, and 0.3 points on ImageNet-C. No standard deviations or number of seeds are reported, and the table cells show single runs. These differences are well within typical run-to-run variation for this benchmark class, so the claim that DCFS is state-of-the-art or competitive is not statistically supported. Please report the mean and standard deviation over at least three seeds with different random orders of the corruption stream (or at least the standard protocol), and indicate whether the differences are reproducible.
  4. [Table 4, Section 4.3] The ablation table is internally ambiguous. The rows with three check marks appear twice (28.5 and 19.3) before the full four-check row, so it is not clear which components are active in each row; the text says 'three modules work together,' but the table suggests four incremental steps. Because this ablation is the only evidence for the contribution of LSCL and the interaction between modules, please rewrite the table so that each row specifies exactly which of LFDC, LCDM, and LSCL is added, and report per-corruption errors as well.
minor comments (6)
  1. [Section 3.2, Eq. (9)] The formula dist(PS,PD) = ||u,v||1 is a typo; it should be ||u − v||1. Also specify what W1 and W2 refer to (e.g., last-layer weight matrices of the two classifiers).
  2. [Section 4.1] The values of mixup α, EMA momentum m, and the exact schedule for μt/σt are not given; please list all hyperparameters in the implementation details.
  3. [Section 3.3] The confidence weighting updates μ and σ with EMA, but the order of operations for the first batch and the handling of μ0=1/C (which is much smaller than typical max(y)) are unclear; please explain.
  4. [Table 3] Per-corruption numbers for AR-TTA and ECoTTA are missing, and the LAW entry '49.08' appears to be a typo (should be '49.0' or '49.8'). Please complete the table.
  5. [Throughout] The paper contains numerous grammatical errors (e.g., 'a adaptive threshold,' 'an initialized,' 'features is represented'), and the claims in the abstract are stronger than the results justify given the ties with PALM; please phrase carefully.
  6. [Figure 2] The sensitivity results are reported only as plots without numeric ranges; state the error rate range for the considered λ values.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: DCFS is a heuristic self-training combination evaluated on external benchmarks; its unvalidated semantic/domain assumption is a validity concern, not a circular derivation.

full rationale

I walked the paper's claimed derivation chain (Eqs. 1–16). Each loss is a self-training or consistency heuristic whose targets are the model's own predictions or hand-defined mixed labels; none of the reported benchmark errors is derived from a parameter fitted to those same errors, and no constant is defined in terms of the result it is said to predict. The reference list contains no prior work by Yin, Sun, Meng, Liu, or Zhou, so there is no self-citation chain carrying the argument. The most vulnerable step is Eq. (1), where a coordinate-attention mask A(F) defines 'semantic-related' and 'domain-related' features as complementary masks; the paper never validates that this split corresponds to semantic versus domain content, and a degenerate mask would still produce well-defined losses. However, that is an unverified modeling assumption and an experimental-validity criticism, not circularity: the paper's equations do not force the claim that the mask captures semantics, and the method is compared against external baselines on CIFAR10-C, CIFAR100-C, and ImageNet-C. The internal ambiguity that P = PS + PD in Eq. (3) may not be normalized is a correctness/implementation issue rather than a circular reduction. Accordingly no circular step meets the evidentiary bar.

Assumptions & free parameters 6 free parameters · 4 assumptions · 1 invented entities

The method relies on an unverified representation decomposition, several hand-chosen loss weights and Gaussian initializations, and an unvalidated confidence proxy. No new physical entities are introduced; the semantic and domain feature split is a conceptual construct without independent evidence.

free parameters (6)
  • lambda_CDM and lambda_SCL = 1.0
    Manual trade-off factors in the total loss (Eq. 16). Sensitivity is shown in Fig. 2, but the reported results use these specific hand-chosen values.
  • lambda in classifier discrepancy term = 0.1
    Weight of the parameter-disagreement regularizer in Eq. (9). The text says it is set to the default value 0.1 without a principled derivation.
  • EMA momentum m = not specified
    Momentum for updating mu_t and sigma_t in Eq. (13) is never given a value, making exact reproduction impossible.
  • Initial Gaussian parameters mu0 and sigma0 = mu0=1/C, sigma0=1.0
    Initialization for the adaptive confidence-weight distribution in Section 3.3, chosen by hand and not derived.
  • lambda_max = 1.0
    Maximum weight in the truncated-Gaussian weighting function in Eq. (11), set manually.
  • Mixup alpha = not specified
    Beta distribution parameter for rho in Eq. (4) is left as alpha in (0, inf); the value used in experiments is not reported.
assumptions (4)
  • domain assumption Target features decompose as FS = F*A(F) and FD = F*(1-A(F)), where A(F) from coordinate attention separates semantic and domain information.
    Section 3.1, Eq. (1): the entire method relies on this decomposition being real; no evidence is given that A(F) tracks semantics versus domain.
  • domain assumption A frozen source-domain classifier h_D encodes domain-related information, while an updated h_S learns target semantic information.
    Section 3.1, Eq. (2): if both classifiers learn the same thing, the dual-classifier design is vacuous.
  • domain assumption The model's maximum softmax probability, after uniform alignment, is a reliable proxy for pseudo-label correctness.
    Section 3.3, Eqs. (11)-(14): the truncated Gaussian weighting and adaptive threshold inherit this assumption; if max confidence is miscalibrated, the weighting can reinforce errors.
  • domain assumption Consistency between sub-feature predictions and whole-feature predictions improves target generalization.
    Section 3.1, Eqs. (3) and (7): this is a modeling bet; the loss can in principle be minimized by output collapse, and no experiment checks that the consistency is informative.
invented entities (1)
  • Semantic-related feature FS and domain-related feature FD
    purpose: Enable dual classifiers to specialize on semantic and domain content respectively, and to support consistency losses between sub-features and whole features.
    The split is generated by a coordinate-attention mask A(F) in Eq. (1), but no external benchmark or falsifiable prediction validates that FS is semantic or FD is domain-related.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DCFS: Continual Test-Time Adaptation via Dual Consistency of Feature and Sample." pith.science (2026). https://pith.science/paper/FEYJUA6A

@misc{pith2026250820516,
  author       = {Pith},
  title        = {Pith review of: DCFS: Continual Test-Time Adaptation via Dual Consistency of Feature and Sample},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FEYJUA6A}},
  note         = {Machine review of arXiv:2508.20516}
}
read the original abstract

Continual test-time adaptation aims to continuously adapt a pre-trained model to a stream of target domain data without accessing source data. Without access to source domain data, the model focuses solely on the feature characteristics of the target data. Relying exclusively on these features can lead to confusion and introduce learning biases. Currently, many existing methods generate pseudo-labels via model predictions. However, the quality of pseudo-labels cannot be guaranteed and the problem of error accumulation must be solved. To address these challenges, we propose DCFS, a novel CTTA framework that introduces dual-path feature consistency and confidence-aware sample learning. This framework disentangles the whole feature representation of the target data into semantic-related feature and domain-related feature using dual classifiers to learn distinct feature representations. By maintaining consistency between the sub-features and the whole feature, the model can comprehensively capture data features from multiple perspectives. Additionally, to ensure that the whole feature information of the target domain samples is not overlooked, we set a adaptive threshold and calculate a confidence score for each sample to carry out loss weighted self-supervised learning, effectively reducing the noise of pseudo-labels and alleviating the problem of error accumulation. The efficacy of our proposed method is validated through extensive experimentation across various datasets, including CIFAR10-C, CIFAR100-C, and ImageNet-C, demonstrating consistent performance in continual test-time adaptation scenarios.

Figures

Figures reproduced from arXiv: 2508.20516 by the authors.

Figure 1
Figure 1. The framework of our proposed DCFS method. [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Results of parameter λCDM(left) and λSCL(right) sensitivity [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 12 canonical work pages

  1. [1]

    arXiv preprint arXiv:2304.04795 (2023) 12 W

    Alfarra, M., Itani, H., Pardo, A., Alhuwaider, S., Ramazanova, M., P´ erez, J.C., Cai, Z., M¨ uller, M., Ghanem, B.: Revisiting test time adaptation under online evaluation. arXiv preprint arXiv:2304.04795 (2023) 12 W. Yin et al

    arXiv 2023

  2. [2]

    In: Proceedings of the AAAI conference on artificial intelligence

    Chen, C., Fu, Z., Chen, Z., Jin, S., Cheng, Z., Jin, X., Hua, X.S.: Homm: Higher- order moment matching for unsupervised domain adaptation. In: Proceedings of the AAAI conference on artificial intelligence. vol. 34, pp. 3422–3429 (2020)

    work page 2020

  3. [3]

    arXiv preprint arXiv:2301.10921 (2023)

    Chen, H., Tao, R., Fan, Y., Wang, Y., Wang, J., Schiele, B., Xie, X., Raj, B., Sav- vides, M.: Softmatch: Addressing the quantity-quality trade-off in semi-supervised learning. arXiv preprint arXiv:2301.10921 (2023)

    arXiv 2023

  4. [4]

    In: European Con- ference on Computer Vision

    Choi, S., Yang, S., Choi, S., Yun, S.: Improving test-time adaptation via shift- agnostic weight regularization and nearest source prototypes. In: European Con- ference on Computer Vision. pp. 440–458. Springer (2022)

    2022

  5. [5]

    arXiv preprint arXiv:2010.09670 (2020)

    Croce, F., Andriushchenko, M., Sehwag, V., Debenedetti, E., Flammarion, N., Chi- ang, M., Mittal, P., Hein, M.: Robustbench: a standardized adversarial robustness benchmark. arXiv preprint arXiv:2010.09670 (2020)

    arXiv 2020

  6. [6]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    D¨ obler, M., Marsden, R.A., Yang, B.: Robust mean teacher for continual and gradual test-time adaptation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 7704–7714 (2023)

    work page 2023

  7. [7]

    Journal of machine learning research 17(59), 1–35 (2016)

    Ganin, Y., Ustinova, E., Ajakan, H., Germain, P., Larochelle, H., Laviolette, F., March, M., Lempitsky, V.: Domain-adversarial training of neural networks. Journal of machine learning research 17(59), 1–35 (2016)

    2016

  8. [8]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Hou, Q., Zhou, D., Feng, J.: Coordinate attention for efficient mobile network de- sign. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 13713–13722 (2021)

    2021

Show all 28 references
  1. [9]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Kang, G., Jiang, L., Yang, Y., Hauptmann, A.G.: Contrastive adaptation network for unsupervised domain adaptation. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 4893–4902 (2019)

    2019

  2. [10]

    Proceedings of the national academy of sciences 114(13), 3521–3526 (2017)

    Kirkpatrick, J., Pascanu, R., Rabinowitz, N., Veness, J., Desjardins, G., Rusu, A.A., Milan, K., Quan, J., Ramalho, T., Grabska-Barwinska, A., et al.: Overcoming catastrophic forgetting in neural networks. Proceedings of the national academy of sciences 114(13), 3521–3526 (2017)

    2017

  3. [11]

    arXiv preprint arXiv:2403.10650 (2024)

    Maharana, S.K., Zhang, B., Guo, Y.: Palm: Pushing adaptive learning rate mech- anisms for continual test-time adaptation. arXiv preprint arXiv:2403.10650 (2024)

    2024 arXiv

  4. [12]

    In: 2022 International Joint Conference on Neural Networks (IJCNN)

    Marsden, R.A., Bartler, A., D¨ obler, M., Yang, B.: Contrastive learning and self- training for unsupervised domain adaptation in semantic segmentation. In: 2022 International Joint Conference on Neural Networks (IJCNN). pp. 1–8. IEEE (2022)

    2022

  5. [13]

    In: International conference on machine learning

    Niu, S., Wu, J., Zhang, Y., Chen, Y., Zheng, S., Zhao, P., Tan, M.: Efficient test- time model adaptation without forgetting. In: International conference on machine learning. pp. 16888–16905. PMLR (2022)

    2022

  6. [14]

    arXiv preprint arXiv:2302.12400 (2023)

    Niu, S., Wu, J., Zhang, Y., Wen, Z., Chen, Y., Zhao, P., Tan, M.: Towards sta- ble test-time adaptation in dynamic wild world. arXiv preprint arXiv:2302.12400 (2023)

    2023 arXiv

  7. [15]

    In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision

    Park, J., Kim, J., Kwon, H., Yoon, I., Sohn, K.: Layer-wise auto-weighting for non-stationary test-time adaptation. In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. pp. 1414–1423 (2024)

    2024

  8. [16]

    Advances in neural information processing systems 33, 11539–11551 (2020)

    Schneider, S., Rusak, E., Eck, L., Bringmann, O., Brendel, W., Bethge, M.: Improv- ing robustness against common corruptions by covariate shift adaptation. Advances in neural information processing systems 33, 11539–11551 (2020)

    2020

  9. [17]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    S´ ojka, D., Cygert, S., Twardowski, B., Trzci´ nski, T.: Ar-tta: A simple method for real-world continual test-time adaptation. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 3491–3495 (2023)

    2023

  10. [18]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Song, J., Lee, J., Kweon, I.S., Choi, S.: Ecotta: Memory-efficient continual test- time adaptation via self-distilled regularization. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 11920–11929 (2023) DCFS: Continual TTA via Feature a...

    2023

  11. [19]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Tzeng, E., Hoffman, J., Saenko, K., Darrell, T.: Adversarial discriminative domain adaptation. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 7167–7176 (2017)

    2017

  12. [20]

    arXiv preprint arXiv:2006.10726 (2020)

    Wang, D., Shelhamer, E., Liu, S., Olshausen, B., Darrell, T.: Tent: Fully test-time adaptation by entropy minimization. arXiv preprint arXiv:2006.10726 (2020)

    2020 arXiv

  13. [21]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Wang, Q., Fink, O., Van Gool, L., Dai, D.: Continual test-time domain adaptation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 7201–7211 (2022)

    2022

  14. [22]

    In: Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing

    Wang, R., Utiyama, M., Liu, L., Chen, K., Sumita, E.: Instance weighting for neural machine translation domain adaptation. In: Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing. pp. 1482–1488 (2017)

    2017

  15. [23]

    In: Proceedings of the 31st ACM International Conference on Multimedia

    Wang, S., Chen, Y., He, Z., Yang, X., Wang, M., You, Q., Zhang, X.: Disentangled representation learning with causality for unsupervised domain adaptation. In: Proceedings of the 31st ACM International Conference on Multimedia. pp. 2918– 2926 (2023)

    2023

  16. [24]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Xie, S., Girshick, R., Doll´ ar, P., Tu, Z., He, K.: Aggregated residual transformations for deep neural networks. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 1492–1500 (2017)

    2017

  17. [25]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Yan, H., Ding, Y., Li, P., Wang, Q., Xu, Y., Zuo, W.: Mind the class weight bias: Weighted maximum mean discrepancy for unsupervised domain adaptation. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 2272–2281 (2017)

    2017

  18. [26]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog- nition

    Yuan, L., Xie, B., Li, S.: Robust test-time adaptation in dynamic scenarios. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog- nition. pp. 15922–15932 (2023)

    2023

  19. [27]

    arXiv preprint arXiv:1605.07146 (2016)

    Zagoruyko, S., Komodakis, N.: Wide residual networks. arXiv preprint arXiv:1605.07146 (2016)

    2016 arXiv

  20. [28]

    arXiv preprint arXiv:1710.09412 (2017)

    Zhang, H., Cisse, M., Dauphin, Y.N., Lopez-Paz, D.: mixup: Beyond empirical risk minimization. arXiv preprint arXiv:1710.09412 (2017)

    2017 arXiv

Pith tools

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