Pith. sign in

REVIEW 4 major objections 5 minor 46 references

Continual Adaptation: Environment-Conditional Parameter Generation for Object Detection in Dynamic Scenarios

T0 review · 4 major / 5 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 A diffusion model that generates detector adapter parameters from the current scene outperforms fine-tuning for continual test-time adaptation.

desk verdict Plausible combination with consistent but modest gains; the central diffusion parameter-generation mechanism is under-specified and needs a clearer description. read the letter →

arxiv 2506.24063 v1 pith:5UT27GPH submitted 2025-06-30 cs.CV

Deng Li , Aming Wu , Yang Li , Yaowei Wang , Yahong Han This is my paper
classification cs.CV
keywords continualtest-timeadaptationobjectdetectionparametergenerationconditionaldiffusionmodelLoRAadapterdomaincatastrophicforgettingoptimaltransport
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

This paper tries to establish that continual test-time adaptation for object detection is better served by generating adapter parameters than by fine-tuning them. The proposed method turns the usual update step into a conditional parameter-generation step: a diffusion model, trained offline on the source domain, receives the current target frame and the trained adapter parameters, and synthesizes robust parameters for a dual-path LoRA adapter. The paper argues this avoids the local optima that unsupervised fine-tuning falls into, while a class-centered optimal transport loss preserves source-domain knowledge. On the Diverse-Weather benchmark the method reports 30.7% average AP versus 29.2% for the SKIP baseline, and 23.0% versus 22.2% on COCO-C with a ResNet-50 backbone.

What carries the argument

The central object is the conditional diffusion-based parameter generator. A latent autoencoder compresses adapter weight matrices into low-dimensional latent vectors, and a diffusion model, conditioned on target-domain image features, denoises a noise vector back to a clean latent that the decoder turns into adapter parameters. Its job is to inject the current environment's scene information into the adapter weights while avoiding the local optima of continual unsupervised gradient updates; the generated parameters are what the detector actually uses for inference.

What would settle it

Run the full pipeline on COCO-C or Diverse-Weather with the diffusion sampler replaced by its mean prediction (one deterministic denoising pass) and with the class-centered optimal transport term replaced by a simple feature-space regularizer; if average AP does not drop meaningfully, the reported gains do not come from the claimed generation and alignment mechanisms. Separately, inspect the implementation of the target-instance to source-class-center assignment: if it needs ground-truth class labels or an oracle assignment, the method is not running as written on an unlabeled stream.

Watch

Extended reading notes

Core claim

The central claim is that replacing the test-time fine-tuning loop with a conditional diffusion-based parameter generator improves continual test-time adaptive object detection. Concretely, the parameter generator takes the current target domain frame and the current adapter parameters as input and generates new adapter parameters via a conditional diffusion model, so that the domain-aware adapter is updated by synthesis rather than by gradient descent on unlabeled target data. Combined with a dual-path LoRA adapter that disentangles features into domain-invariant and domain-specific components, and a class-centered optimal transport alignment that anchors target instances to source class centers, the method reports state-of-the-art results on Diverse-Weather, COCO-C, and SHIFT benchmarks, with the largest gain on Diverse-Weather at 30.7% average AP versus 29.2% for the SKIP baseline.

Load-bearing premise

The method's forgetting-mitigation loss assumes that at test time each unlabeled target image's detected instances can be reliably assigned to the right source class center; the paper does not say how that assignment is computed, filtered, or verified under domain shift.

Editorial extensions

If this is right

  • The generator is frozen at test time and only the lightweight adapter is updated, so adaptation avoids modifying the full backbone and stays efficient for online use.
  • The dual-path LoRA adapter separates domain-invariant from domain-specific features, and the paper's feature-map visualizations show the generated parameters concentrate attention on objects rather than background.
  • Class-centered optimal transport alignment preserves source-domain class structure while adapting, which the paper ties to higher source-domain accuracy after adaptation on COCO (42.4% with ResNet-50).
  • Because adaptation becomes parameter generation, the approach is architecture-agnostic about the detector body; the paper demonstrates it on both CNN (ResNet-50/101) and transformer (Swin-T) backbones.
  • The mechanism is not tied to any particular corruption or weather type, so the same pipeline serves weather changes, synthetic corruptions, and autonomous-driving domain sequences.

Reading between the lines

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

  • If the generation mechanism is what drives the gain, a cheaper deterministic generator (for example a conditional MLP mapping scene features directly to adapter weights) should underperform the diffusion sampler; that comparison would isolate whether stochastic denoising, rather than the conditioning, is the active ingredient.
  • The method implies a memory-forgetting trade-off: the same source class centers used by the optimal transport loss could be refreshed with adapted features, which the paper does not explore but which would test whether alignment is better viewed as anchoring or as a moving target.
  • A testable extension would be to vary the number of reverse diffusion steps at inference; if the gains are robust to aggressive step reduction, the generator might be acting mostly as a regularizer, and the paper's claim about escaping local optima could be reformulated as a claim about parameter-space smoothing.
  • Because the paper leaves the target-instance-to-class assignment under-specified, an oracle-ablation (perfect pseudo-labels versus no pseudo-labels) would reveal how much of the reported gain rests on the undocumented assignment step.
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 / 5 minor

Summary. The paper proposes a continual test-time adaptive object detection (CTTAOD) method that combines a dual-path LoRA-based domain-aware adapter, a conditional diffusion-based parameter generator that outputs adapter weights conditioned on the current target image and current adapter weights, and a class-centered optimal transport alignment loss for mitigating catastrophic forgetting. The method is evaluated on Diverse-Weather, COCO-C, and SHIFT with Faster R-CNN backbones (ResNet-101, ResNet-50, Swin-T), reporting consistent improvements over SKIP and other baselines, with average AP gains of 1.5, 0.8-0.9, and 0.5-1.0 points respectively. The paper also includes ablations and feature-map visualizations.

Significance. If fully specified and reproducible, the idea of generating adapter parameters rather than directly fine-tuning them is a plausible and interesting departure from standard CTTAOD, and the consistent gains across three benchmarks and two backbone families suggest the overall approach is worth pursuing. The paper also explicitly targets the local-optima and catastrophic-forgetting challenges. However, the current manuscript does not provide an executable specification of the central diffusion-based parameter generation, does not describe how pseudo-labels for the OT alignment are obtained, and omits hyperparameters and error bars; these omissions currently prevent verification of the claimed mechanism.

major comments (4)
  1. [§3.4, Eq. (9)] The inference procedure for the conditional diffusion generator Diff(·) is not specified: the paper does not state how the current adapter weights W_Adapter are encoded into the latent z0 or a noised latent, what timestep and noise schedule is used, how many denoising steps are taken, or how the conditioning image x^T_t is encoded. As written, Eq. (9) is not implementable, and the ablation gains attributed to 'Parameter-Gen' in Table 4 cannot be verified or attributed to the stated diffusion mechanism. Moreover, Eq. (8) conditions on x^T during training, but the generator is trained on source-domain data, so it is unclear whether source images are used as conditioning and how this transfers to target images.
  2. [§3.5, Eqs. (10)-(12)] The class-centered OT loss requires target instances assigned to class c (x^c_{t,j}), but the target stream is unlabeled and the paper never specifies how pseudo-labels are generated, filtered, or weighted. Without this, the loss cannot be computed in the reported test-time protocol. In addition, because each source class is represented by a single center μ^c_s, the transport plan in Eq. (11) has a single non-zero entry per instance under the standard marginal constraints, so the OT formulation degenerates to a plain squared-distance loss and the proposed 'optimal transport alignment' adds no optimization over plans.
  3. [§4.2, Eq. (13)] The paper does not report values for the loss weights λ_orth, λ_HSIC, λ_A, and λ_CA, the LoRA ranks r1 and r2, or the diffusion hyperparameters (number of timesteps T, noise schedule, number of sampling steps). Without these, the experiments cannot be reproduced, and it is unknown whether the reported gains are sensitive to these choices.
  4. [§3.4, Table 4] The conditional parameter generator is trained offline on source-domain data only and frozen during test time; for Diverse-Weather the source is Daytime Clear, yet the conditioning images at test time are night, fog, and rain. The paper provides no control experiment (e.g., conditioning on a corrupted or random image, or ablating the conditioning input) to show that the generated parameters respond to the target environment rather than acting as a fixed stochastic perturbation. Without such a probe, the +1.2/+0.6 AP gains attributed to parameter generation in Table 4 are not convincingly explained.
minor comments (5)
  1. [§4.3, SHIFT paragraph] The text states a gain of '1.0% (41.0% - 30.0%)' for SHIFT-Discrete with Swin-T, but Table 3 lists the SKIP value as 40.6, not 30.0; the reported difference should be 0.4%.
  2. [§3.3, Eq. (1)] The dimensions of B_sp are written as R^{r1×d}, but since A_sp is in R^{d×r2}, B_sp should be in R^{r2×d}; the ranks are inconsistent as printed.
  3. [§3.4, final paragraph] 'domain-ware adapters' should read 'domain-aware adapters'.
  4. [Table 2 caption] The corruption abbreviations (Gau, Sht, Imp, Def, Gls, Mtn, Zm, Snw, Frs, Fog, Brt, Cnt, Els, Px, Jpg) are not defined; a caption listing the full names would improve readability.
  5. [Tables 1-3] The tables report single-run numbers without standard deviations; given the small reported margins on some SHIFT rows, error bars or multiple seeds are needed to assess whether the gains are significant.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation found: the parameter-generation, adapter, and alignment losses are defined by explicit equations and evaluated against external baselines; the only self-citations are non-load-bearing.

full rationale

The claimed derivation chain is not circular. The dual-path LoRA adapter is defined by Eq. (1) with orthogonality and HSIC losses (2)-(5), all specified in the paper. The conditional diffusion parameter generator is defined by forward/reverse processes (7)-(8) and inference (9); although Eq. (9) is underspecified (Diff is not given as an executable procedure), underspecification is a reproducibility concern, not a reduction of the output to the input by construction. The class-centered OT alignment (10)-(12) is a stated loss; the lack of pseudo-label details is an implementation gap, not circularity. The only self-citations are [41] (Diverse-Weather dataset and visualization) and [42] (disentanglement related work); neither supplies a load-bearing theorem or fitted value, and the central equations do not reduce to them. Ablation results in Table 4 compare against LoRA and KL-alignment variants, and the method is benchmarked against external methods (SKIP, NORM, DUA, etc.), so the reported gains are not fitted inputs renamed as predictions. Therefore no specific reduction can be quoted, and the score is 0.

Assumptions & free parameters 3 free parameters · 4 assumptions · 2 invented entities

The central method rests on several unreported implementation choices (loss weights, ranks, pseudo-labeling) and on assumptions that the frozen generator transfers to unseen domains and that class labels are available for target instances. These are not derived from first principles.

free parameters (3)
  • Loss weights (λ_orth, λ_HSIC, λ_A, λ_CA) = not reported
    Weights for orthogonality, HSIC, adapter, and class-center alignment losses in Eqs. (5) and (13) are chosen by hand; values are not given in the paper.
  • LoRA ranks r1 and r2 = not reported
    Rank of low-rank matrices A_inv B_inv and A_sp B_sp in Eq. (1); no values given.
  • Diffusion schedule and steps T = not reported
    Noise schedule β_t and number of diffusion steps in Eqs. (7)-(8) are standard but unspecified.
assumptions (4)
  • domain assumption The detector is pre-trained on a labeled source domain, and only the adapter parameters are updated at test time.
    Problem formulation in Section 3.1; standard for CTTAOD.
  • ad hoc to paper A diffusion generator trained offline on source-domain parameter states can produce useful adapter parameters for unseen target domains when conditioned on target images.
    Section 3.4 claims the generator 'learns robust parameter generation capabilities' from source data and is then frozen; this generalization is assumed but not analyzed.
  • ad hoc to paper Aligning target instance features to fixed source class centers via optimal transport mitigates catastrophic forgetting.
    Section 3.5 makes this claim without evaluating forgetting beyond a few source-domain AP numbers.
  • domain assumption Target instances can be assigned to class labels at test time for the OT alignment.
    Section 3.5 defines {x_{c,t,j}} as target instances 'in class c' without specifying how class labels are obtained.
invented entities (2)
  • Conditional diffusion-based parameter generator
    purpose: Synthesizes LoRA adapter parameters conditioned on target scene features, intended to prevent local optima.
    New module introduced in Section 3.4; no independent evidence beyond the paper's own experiments.
  • Dual-path LoRA domain-aware adapter
    purpose: Disentangles features into domain-invariant and domain-specific components in the backbone.
    New adapter introduced in Section 3.3; no independent validation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Continual Adaptation: Environment-Conditional Parameter Generation for Object Detection in Dynamic Scenarios." pith.science (2026). https://pith.science/paper/5UT27GPH

@misc{pith2026250624063,
  author       = {Pith},
  title        = {Pith review of: Continual Adaptation: Environment-Conditional Parameter Generation for Object Detection in Dynamic Scenarios},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5UT27GPH}},
  note         = {Machine review of arXiv:2506.24063}
}
read the original abstract

In practice, environments constantly change over time and space, posing significant challenges for object detectors trained based on a closed-set assumption, i.e., training and test data share the same distribution. To this end, continual test-time adaptation has attracted much attention, aiming to improve detectors' generalization by fine-tuning a few specific parameters, e.g., BatchNorm layers. However, based on a small number of test images, fine-tuning certain parameters may affect the representation ability of other fixed parameters, leading to performance degradation. Instead, we explore a new mechanism, i.e., converting the fine-tuning process to a specific-parameter generation. Particularly, we first design a dual-path LoRA-based domain-aware adapter that disentangles features into domain-invariant and domain-specific components, enabling efficient adaptation. Additionally, a conditional diffusion-based parameter generation mechanism is presented to synthesize the adapter's parameters based on the current environment, preventing the optimization from getting stuck in local optima. Finally, we propose a class-centered optimal transport alignment method to mitigate catastrophic forgetting. Extensive experiments conducted on various continuous domain adaptive object detection tasks demonstrate the effectiveness. Meanwhile, visualization results show that the representation extracted by the generated parameters can capture more object-related information and strengthen the generalization ability.

Figures

Figures reproduced from arXiv: 2506.24063 by the authors.

Figure 1
Figure 1. Illustration of proposed parameter generation method for object detector in continually changing scenarios. The conditional parameter generator generates robust parameters for the detector with the target scene conditions to improve the detector’s generalization performance. scenarios. In practical scenarios, there is a distribution dif￾ference between the training data (source domain) and the test data (target doma… view at source ↗
Figure 2
Figure 2. Illustration of our proposed parameter-generated adapter for continual test-time adaptive object detection. This method mainly includes the conditional parameter generator, domain-aware adapter, and class-centered optimal transport alignment module. The domain-aware adapter is introduced into multiple blocks of the backbone. It disentangles the features into domain-invariant features Finv and domain-specific feature… view at source ↗
Figure 3
Figure 3. Detection results of the continual test-time adaptation on the urban scene Diverse-Weather Dataset [41], where the top row represents the detection results of SKIP [43], and the bottom row corresponds to our proposed method. In the “Daytime Foggy” scene, our method accurately detects small-sized cars. In the “Dusk Rainy” scene, SKIP [43] incorrectly detects an object as a bus. In the complex “Night Rainy” scene, SKI… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Visualization results of feature maps on Diverse-Weather Dataset [41]. The top row is the input image, the second row is the results of the baseline method SKIP [43], and the bottom row is the results of our proposed method. Method Adapter Parameter-Gen Align Diverse-W…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

46 extracted references · 39 canonical work pages

  1. [1]

    Hyper- style: Stylegan inversion with hypernetworks for real image editing

    Yuval Alaluf, Or Patashnik, and Daniel Cohen-Or. Hyper- style: Stylegan inversion with hypernetworks for real image editing. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022. 3

    work page 2022

  2. [2]

    A theory of learning from different domains.Machine learn- ing, 79:151–175, 2010

    Shai Ben-David, John Blitzer, Koby Crammer, Alex Kulesza, Fernando Pereira, and Jennifer Wortman Vaughan. A theory of learning from different domains.Machine learn- ing, 79:151–175, 2010. 2

    work page 2010

  3. [3]

    End-to- end object detection with transformers

    Nicolas Carion, Francisco Massa, Gabriel Synnaeve, Nicolas Usunier, Alexander Kirillov, and Sergey Zagoruyko. End-to- end object detection with transformers. In European confer- ence on computer vision, pages 213–229. Springer, 2020. 1

    2020

  4. [4]

    Domain generalization by solving jigsaw puzzles

    Fabio M Carlucci, Antonio D’Innocente, Silvia Bucci, Bar- bara Caputo, and Tatiana Tommasi. Domain generalization by solving jigsaw puzzles. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 2229–2238, 2019. 1

    work page 2019

  5. [5]

    Stfar: Im- proving object detection robustness at test-time by self- training with feature alignment regularization.arXiv preprint arXiv:2303.17937, 2023

    Yijin Chen, Xun Xu, Yongyi Su, and Kui Jia. Stfar: Im- proving object detection robustness at test-time by self- training with feature alignment regularization.arXiv preprint arXiv:2303.17937, 2023. 2

    arXiv 2023

  6. [6]

    Robustnet: Improving domain generalization in urban-scene segmentation via in- stance selective whitening

    Sungha Choi, Sanghun Jung, Huiwon Yun, Joanne T Kim, Seungryong Kim, and Jaegul Choo. Robustnet: Improving domain generalization in urban-scene segmentation via in- stance selective whitening. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 11580–11590, 2021. 6

    work page 2021

  7. [7]

    Hyperdiffusion: Diffusion models for neural im- plicit fields via hypernetworks

    Mehmet Erkoc, Ozan Kar, Caglar Budak, Riad Akrour, Emre Sahin, et al. Hyperdiffusion: Diffusion models for neural im- plicit fields via hypernetworks. In International Conference on Learning Representations (ICLR), 2023. 3

    work page 2023

  8. [8]

    Cloud-device col- laborative adaptation to continual changing environments in the real-world

    Yulu Gan, Mingjie Pan, Rongyu Zhang, Zijian Ling, Lingran Zhao, Jiaming Liu, and Shanghang Zhang. Cloud-device col- laborative adaptation to continual changing environments in the real-world. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 12157– 12166, 2023. 2, 3

    work page 2023

Show all 46 references
  1. [9]

    Unsupervised domain adaptation by backpropagation

    Yaroslav Ganin and Victor Lempitsky. Unsupervised domain adaptation by backpropagation. In International conference on machine learning, pages 1180–1189. PMLR, 2015. 1

    2015

  2. [10]

    Domain-adversarial train- ing of neural networks

    Yaroslav Ganin, Evgeniya Ustinova, Hana Ajakan, Pas- cal Germain, Hugo Larochelle, Francois Laviolette, Mario Marchand, and Victor Lempitsky. Domain-adversarial train- ing of neural networks. In International Conference on Ma- chine Learning, pages 2093–2102. PMLR, 2016. 2

    2016

  3. [11]

    Domain generalization for object recog- nition with multi-task autoencoders

    Muhammad Ghifary, W Bastiaan Kleijn, Mengjie Zhang, and David Balduzzi. Domain generalization for object recog- nition with multi-task autoencoders. In Proceedings of the IEEE international conference on computer vision , pages 2551–2559, 2015. 1

    2015

  4. [12]

    Hypernetworks

    David Ha, Andrew Dai, and Quoc V Le. Hypernetworks. In International Conference on Learning Representations (ICLR), 2017. 3

    2017

  5. [13]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 770–778, 2016. 1, 6

    2016

  6. [14]

    Lora: Low-rank adaptation of large language models

    Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen- Zhu, Yuanzhi Li, Shean Wang, Lu Wang, Weizhu Chen, et al. Lora: Low-rank adaptation of large language models. ICLR, 1(2):3, 2022. 3, 8

    2022

  7. [15]

    It- erative normalization: Beyond standardization towards effi- cient whitening

    Lei Huang, Yi Zhou, Fan Zhu, Li Liu, and Ling Shao. It- erative normalization: Beyond standardization towards effi- cient whitening. In Proceedings of the ieee/cvf conference on computer vision and pattern recognition, pages 4874–4883,

  8. [16]

    Overcoming catastrophic forgetting in neu- ral networks

    James Kirkpatrick, Razvan Pascanu, Neil Rabinowitz, Joel Veness, Guillaume Desjardins, Andrei A Rusu, Antonio Milan, John Quan, Tiago Ramalho, Agnieszka Grabska- Barwinska, et al. Overcoming catastrophic forgetting in neu- ral networks. Proceedings of the national academy of s...

    2017

  9. [17]

    Universal source-free domain adaptation

    Jogendra Nath Kundu, Naveen Venkat, R Venkatesh Babu, et al. Universal source-free domain adaptation. In Proceed- ings of the IEEE/CVF conference on computer vision and pattern recognition, pages 4544–4553, 2020. 1

    2020

  10. [18]

    Microsoft coco: Common objects in context

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Doll´ar, and C Lawrence Zitnick. Microsoft coco: Common objects in context. In European conference on computer vision , pages 740–755. Springer, 2014. 5

    2014

  11. [19]

    Bird’s-eye-view scene graph for vision-language navigation

    Rui Liu, Xiaohan Wang, Wenguan Wang, and Yi Yang. Bird’s-eye-view scene graph for vision-language navigation. In ICCV, 2023. 1

    2023

  12. [20]

    Vision-language nav- igation with energy-based policy

    Rui Liu, Wenguan Wang, and Yi Yang. Vision-language nav- igation with energy-based policy. In NeurIPS, 2024

    2024

  13. [21]

    V olumetric envi- ronment representation for vision-language navigation

    Rui Liu, Wenguan Wang, and Yi Yang. V olumetric envi- ronment representation for vision-language navigation. In CVPR, 2024. 1

    2024

  14. [22]

    Swin transformer: Hierarchical vision transformer using shifted windows

    Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, and Baining Guo. Swin transformer: Hierarchical vision transformer using shifted windows. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 10012–10022, 2021. 6

    2021

  15. [23]

    Benchmarking ro- bustness in object detection: Autonomous driving when win- ter is coming

    Claudio Michaelis, Benjamin Mitzkus, Robert Geirhos, Evgenia Rusak, Oliver Bringmann, Alexander S Ecker, Matthias Bethge, and Wieland Brendel. Benchmarking ro- bustness in object detection: Autonomous driving when win- ter is coming. arXiv preprint arXiv:1907.07484, 2019. 6

    1907 arXiv

  16. [24]

    The norm must go on: Dynamic unsuper- vised domain adaptation by normalization

    M Jehanzeb Mirza, Jakub Micorek, Horst Possegger, and Horst Bischof. The norm must go on: Dynamic unsuper- vised domain adaptation by normalization. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 14765–14775, 2022. 2, 6

    2022

  17. [25]

    Act- mad: Activation matching to align distributions for test- time-training

    Muhammad Jehanzeb Mirza, Pol Jan ´e Soneira, Wei Lin, Ma- teusz Kozinski, Horst Possegger, and Horst Bischof. Act- mad: Activation matching to align distributions for test- time-training. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pag...

    2023

  18. [26]

    Domain generalization via invariant fea- ture representation

    Krikamol Muandet, David Balduzzi, and Bernhard Sch¨olkopf. Domain generalization via invariant fea- ture representation. In International conference on machine learning, pages 10–18. PMLR, 2013. 1

    2013

  19. [27]

    Image to image transla- tion for domain adaptation

    Zak Murez, Soheil Kolouri, David Kriegman, Ravi Ra- mamoorthi, and Kyungnam Kim. Image to image transla- tion for domain adaptation. In Proceedings of the IEEE con- ference on computer vision and pattern recognition , pages 4500–4509, 2018. 1

    2018

  20. [28]

    Towards stable test-time adaptation in dynamic wild world

    Shuaicheng Niu, Jiaxiang Wu, Yifan Zhang, Zhiquan Wen, Yaofo Chen, Peilin Zhao, and Mingkui Tan. Towards stable test-time adaptation in dynamic wild world. arXiv preprint arXiv:2302.12400, 2023. 2

    2023 arXiv

  21. [29]

    Two at once: Enhancing learning and generalization capacities via ibn-net

    Xingang Pan, Ping Luo, Jianping Shi, and Xiaoou Tang. Two at once: Enhancing learning and generalization capacities via ibn-net. In Proceedings of the European Conference on Computer Vision (ECCV), pages 464–479, 2018. 6

    2018

  22. [30]

    Switchable whitening for deep representa- tion learning

    Xingang Pan, Xiaohang Zhan, Jianping Shi, Xiaoou Tang, and Ping Luo. Switchable whitening for deep representa- tion learning. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 1863–1871, 2019. 6

    2019

  23. [31]

    Learning to learn with genera- tive models of neural network checkpoints

    William Peebles, Ilija Radosavovic, Tim Brooks, Alexei A Efros, and Jitendra Malik. Learning to learn with genera- tive models of neural network checkpoints. arXiv preprint arXiv:2209.12892, 2022. 3

    2022 arXiv

  24. [32]

    Deep convolutional neu- ral networks for image classification: A comprehensive re- view

    Waseem Rawat and Zenghui Wang. Deep convolutional neu- ral networks for image classification: A comprehensive re- view. Neural computation, 29(9):2352–2449, 2017. 1

    2017

  25. [33]

    Fully test-time adaptation for object detection

    Xiaoqian Ruan and Wei Tang. Fully test-time adaptation for object detection. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 1038–1047, 2024. 2

    2024

  26. [34]

    Generate to adapt: Aligning domains using generative adversarial networks

    Swami Sankaranarayanan, Yogesh Balaji, Carlos D Castillo, and Rama Chellappa. Generate to adapt: Aligning domains using generative adversarial networks. In Proceedings of the IEEE conference on computer vision and pattern recogni- tion, pages 8503–8512, 2018. 1

    2018

  27. [35]

    Improving robustness against common corruptions by covariate shift adaptation

    Steffen Schneider, Evgenia Rusak, Luisa Eck, Oliver Bring- mann, Wieland Brendel, and Matthias Bethge. Improving robustness against common corruptions by covariate shift adaptation. Advances in neural information processing sys- tems, 33:11539–11551, 2020. 2, 6

    2020

  28. [36]

    Ecotta: Memory-efficient continual test-time adaptation via self-distilled regularization

    Junha Song, Jungsoo Lee, In So Kweon, and Sungha Choi. Ecotta: Memory-efficient continual test-time adaptation via self-distilled regularization. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 11920–11929, 2023. 2

    2023

  29. [37]

    Mixture regression for covariate shift

    Masashi Sugiyama and Amos J Storkey. Mixture regression for covariate shift. Advances in neural information process- ing systems, 19, 2006. 1

    2006

  30. [38]

    Shift: a synthetic driving dataset for continuous multi-task domain adaptation

    Tao Sun, Mattia Segu, Janis Postels, Yuxuan Wang, Luc Van Gool, Bernt Schiele, Federico Tombari, and Fisher Yu. Shift: a synthetic driving dataset for continuous multi-task domain adaptation. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognitio...

    2022

  31. [39]

    Tent: Fully test-time adaptation by entropy minimization

    Dequan Wang, Evan Shelhamer, Shaoteng Liu, Bruno Ol- shausen, and Trevor Darrell. Tent: Fully test-time adaptation by entropy minimization. In International Conference on Learning Representations, 2020. 1, 2

    2020

  32. [40]

    Continual test-time adaptation

    Dequan Wang, Evan Shelhamer, Shaoteng Liu, Trevor Dar- rell, and Alexei A Efros. Continual test-time adaptation. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 7201–7211, 2022. 2

    2022

  33. [41]

    Single-domain generalized ob- ject detection in urban scene via cyclic-disentangled self- distillation

    Aming Wu and Cheng Deng. Single-domain generalized ob- ject detection in urban scene via cyclic-disentangled self- distillation. In Proceedings of the IEEE/CVF Conference on computer vision and pattern recognition, pages 847–856,

  34. [42]

    Vector-decomposed disentanglement for domain- invariant object detection

    Aming Wu, Rui Liu, Yahong Han, Linchao Zhu, and Yi Yang. Vector-decomposed disentanglement for domain- invariant object detection. In ICCV, pages 9342–9351, 2021. 1

    2021

  35. [43]

    Jayeon Yoo, Dongkwan Lee, Inseop Chung, Donghyun Kim, and Nojun Kwak. What how and when should object detec- tors update in continually changing test domains? In Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 23354–23363, 2024. 2, 3, 6, 7, 8

    2024

  36. [44]

    Spatio-temporal few-shot learning via diffusive neural network generation

    Yuan Yuan, Chenyang Shao, Jingtao Ding, Depeng Jin, and Yong Li. Spatio-temporal few-shot learning via diffusive neural network generation. In The Twelfth International Conference on Learning Representations, 2024. 3

    2024

  37. [45]

    Memo: Test time robustness via adaptation and augmentation

    Marvin Zhang, Sergey Levine, and Chelsea Finn. Memo: Test time robustness via adaptation and augmentation. Ad- vances in neural information processing systems, 35:38629– 38642, 2022. 1, 2

    2022

  38. [46]

    Object detection in 20 years: A survey.Proceed- ings of the IEEE, 2023

    Zhengxia Zou, Keyan Chen, Zhenwei Shi, Yuhong Guo, and Jieping Ye. Object detection in 20 years: A survey.Proceed- ings of the IEEE, 2023. 1

    2023

Pith tools

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