Pith. sign in

REVIEW 4 major objections 4 minor 66 references

SS-DC: Spatial-Spectral Decoupling and Coupling Across Visible-Infrared Gap for Domain Adaptive Object Detection

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

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

Pith's one-line read This paper claims that decoupling RGB source imagery into domain-invariant and domain-specific spectral components, then coupling them with spatial features, closes much of the visible-to-infrared domain gap in unsupervised object…

desk verdict A plausible spectral decoupling framework for RGB-IR UDAOD, but the core idempotence and decoupling claims don't survive contact with the equations. read the letter →

arxiv 2507.12017 v1 pith:3GDSFAZB submitted 2025-07-16 cs.CV cs.AI

Xiwei Zhang , Chunjin Yang , Yiming Xiao , Runtong Zhang , Fanman Meng This is my paper · ORCID
classification cs.CVcs.AI
keywords unsuperviseddomainadaptiveobjectdetectionRGB-infraredadaptationspectraldecouplingfrequencyfilteringmeanteachertransformerdomain-invariantfeaturesFLIR-ADAS
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 show that a major obstacle in adapting object detectors from visible (RGB) to infrared (IR) images is treating the RGB source as one uniform domain, when it actually contains subdomains such as daytime, nighttime, and fog. It proposes a framework, SS-DC, that splits each image's frequency spectrum into domain-invariant (DI) and domain-specific (DS) content, couples the DI features with spatial features at multiple scales, and feeds DS tokens into the transformer encoder. The authors report that on the FLIR-ADAS protocol, SS-DC reaches 48.02 mAP versus 43.04 for the previous state-of-the-art D3T, and on FLIR, 72.43 versus 69.30. The claim matters because it suggests spectral decoupling of a heterogeneous RGB source is a usable lever for RGB-to-IR adaptation without forcing direct feature alignment.

What carries the argument

The load-bearing mechanism is the Spectral Adaptive Idempotent Decoupling (SAID) module together with the Spatial-Spectral Coupling (SS-Coupling) strategy. SAID works in the 2D Fourier domain: input images are converted to amplitude and phase spectra, then processed by a filter bank (soft mode), Gaussian filters (hard mode), or a lightweight U-Net (free mode) to create an adaptive filter $H_{inv}$ for DI features and its complement $H_{spe}$ for DS features. A self-distillation decoupling loss, Eq. 5, feeds DI through a parameter-shared second SAID and uses Pearson correlation to require DI to stay correlated across passes while DS decorrelates, the mechanism intended to enforce idempotence. SS-Coupling then embeds the decoupled spectral features into the spatial backbone: shallow layers get attention-map-guided feature selection, deep layers get cross-attention, and DS tokens are injected into the DETR encoder to suppress domain-specific interference.

What would settle it

Compare SS-DC against the same architecture with the decoupling loss replaced by a matched-strength auxiliary loss that does not separate DI and DS, such as random filter regularization; if mAP stays near 48.02, the spectral-decoupling mechanism is not the cause. Alternatively, on the IR test set, feed each image through SAID twice and check whether the Pearson correlation between the two DI outputs is close to 1 while DS outputs decorrelate; failure would show the idempotence premise is not met.

Watch

Extended reading notes

Core claim

The central discovery claimed is that explicit spectral decoupling improves RGB-to-infrared unsupervised domain adaptive object detection. The paper's SAID module applies adaptive filters in the Fourier domain, with hard, soft, and free modes, to decompose amplitude and phase spectra into DI and DS parts; a self-distillation loss based on the Pearson correlation coefficient keeps repeated filtering of DI highly correlated while forcing DS artifacts to decorrelate. The SS-Coupling stage then merges spectral DI features with spatial backbone features through attention-map-guided blending in early layers and cross-attention in later layers, while DS tokens guide the DETR encoder to suppress domain bias. On the new FLIR-ADAS protocol, with 5,663 labeled RGB images, 4,856 unlabeled IR images, and 1,144 IR test images, SS-DC obtains 48.02 mAP, and 72.43 mAP on FLIR, outperforming D3T. The authors interpret this as evidence that modeling the multi-subdomain structure of the RGB source benefits adaptation to IR.

Load-bearing premise

The paper's gains depend on the assumption that the soft self-distillation loss truly separates domain-invariant from domain-specific spectral content in a way that transfers to infrared, rather than simply providing an additional strong training signal.

Editorial extensions

If this is right

  • If the reported gains hold, treating the RGB source as a mixture of subdomains and explicitly separating DI and DS content becomes a viable alternative to direct feature alignment for RGB-IR unsupervised domain adaptation.
  • The 48.02 mAP on FLIR-ADAS and 72.43 mAP on FLIR suggest the method transfers to both heterogeneous (multi-weather and glare) and relatively uniform source-domain settings.
  • The soft filter-bank mode outperforming hard and free modes indicates that adaptive frequency partitioning is more effective than fixed Gaussian filters or fully free learned filters.
  • Introducing DS information as guidance tokens into the DETR encoder is claimed to reduce domain bias without destabilizing detection loss convergence.

Reading between the lines

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

  • Editorial extension: the paper leaves implicit that a matched-strength control experiment, replacing the decoupling loss with a non-spectral auxiliary loss such as random filter regularization, would isolate whether the gain comes from true DI/DS separation or merely from extra training signal.
  • Editorial extension: the idempotence property is only softly enforced, so one could directly test on IR images whether repeated SAID passes reach a fixed point and whether DI components transfer between RGB subdomains and IR better than DS components.
  • Editorial extension: the approach may generalize to other large-gap adaptation problems with heterogeneous sources, such as SAR-to-optical or synthetic-to-real detection, although the paper does not test this.
  • Editorial extension: an explicit DI/DS correlation diagnostic on the target domain would turn the qualitative claim of domain-bias reduction into a measurable quantity.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper proposes SS-DC, a mean-teacher/DETR framework for unsupervised domain adaptive object detection from RGB to infrared. It introduces SAID, a spectral filtering module with hard/soft/free modes intended to decompose amplitude/phase spectra into domain-invariant (DI) and domain-specific (DS) components, trained with a Pearson-correlation self-distillation loss (Eq. 5). SS-Coupling then integrates the spectral DI features with spatial backbone features and injects DS tokens into the DETR encoder. The paper reports 48.02 mAP on a new FLIR-ADAS RGB-to-IR protocol (versus 43.04 for D3T) and 72.43 mAP on FLIR (versus 69.30 for D3T), with ablations attributing the gains to SAID, SS-Coupling, and SSM.

Significance. If the central decoupling mechanism is real, the paper is useful: it broadens UDAOD to multi-subdomain RGB sources, and the new FLIR-ADAS protocol addresses a genuine weakness of previously used aligned FLIR benchmarks. The empirical gains are large, and the ablation sequence in Table 3 is monotonic and internally consistent. However, the paper does not yet establish that spectral decoupling is the cause of the gains: the proposed loss does not mathematically enforce idempotence or DI/DS separation, and the reported best mAP is selected from a grid search with no error bars. These issues are load-bearing for the paper's central claim and must be resolved before the results can be accepted as evidence for the proposed mechanism.

major comments (4)
  1. [Sec. 3.4, Eq. (5)] Equation (5) does not enforce idempotence or DI/DS separation, and it admits a trivial solution: if the learned filter is the identity for the invariant branch (H_inv=1, H_spe=0), then DS=DS'=DS''=0, making the numerator zero and the denominator one, so L_dcp is minimized while no decoupling occurs. In addition, the hard-mode filter F_hard(A)=A*(1-exp(-D^2/(2*sigma_h^2))) is not idempotent: applying it twice multiplies by (1-exp(-D^2/(2*sigma_h^2)))^2, so the cited Semi-Group Property of Gaussian functions does not imply F^n=F for this pointwise mask. The loss only compares correlations across three successive passes; it never compares DS with a domain or subdomain label, nor DI with semantic content across source subdomains and the IR target, so the labels 'domain-invariant' and 'domain-specific' are not grounded. The row labeled 'SAID' in Table 3 removes the whole module, so the observed +3.91 mAP gain could come from added parameters or the extra training signal rather than from genuine spectral decoupling.
  2. [Sec. 4.2, Tables 4-6, Fig. 5] The paper states in Section 4.2 that each experiment was repeated three times and averaged, but no variance or per-seed results are reported anywhere, so the reader cannot assess whether the 4.98-point FLIR-ADAS advantage over D3T is statistically meaningful. More seriously, Tables 4-6 and Figure 5 are sweeps over SAID mode, number of filters N, decoupling loss strength k, loss coefficient lambda_dcp, and SSM step size, all evaluated on the same FLIR-ADAS target test set; the reported 48.02 mAP is the maximum over these sweeps (soft mode, N=200, k=2, lambda_dcp=50, SSM step 500). No validation split or model-selection procedure is described, so the headline number is likely an optimistic estimate of the method's expected performance. The authors should report mean +/- standard deviation over their three runs with a configuration fixed on a validation subset, and ideally compare against the baseline under the same selection protocol.
  3. [Sec. 4.1] The new FLIR-ADAS protocol is a contribution of the paper and is load-bearing for all main results, but it is not specified sufficiently for reproduction. Section 4.1 gives a split of 5,663 labeled RGB images, 4,856 unlabeled IR images, and 1,144 IR evaluation images, but it does not describe how this split was created, how 'weak alignment' was eliminated, which of the 15 object categories were retained beyond the five reported, or how the claimed multi-subdomain structure (daytime, nighttime, glare) is represented in the source training set. The text refers the reader to an appendix, but the appendix is not part of the submitted manuscript. These details need to be provided before the protocol can serve as a reproducible benchmark.
  4. [Sec. 3.3, Eq. (3)] The Student Stabilization Module contributes +1.04 mAP in the ablation (Table 3, rows 3-4), but it is only specified by the equation theta_s <- alpha_ssm * theta_s + (1 - alpha_ssm) * theta_t and a citation to prior work. The 'SSM step size' used in Figure 5 is never defined, and the update schedule and interaction with the EMA update are not described. Without this information, the SSM contribution and the final 48.02 mAP result cannot be reproduced.
minor comments (4)
  1. [Throughout] There are several typos and grammatical issues, including 'muti-subdomain' in the Introduction, 'RGH-to-IR' and 'FILR-ADAS' in Section 4.1, 'precious works' instead of 'previous works' in Section 4.4, 'Student Stabilisation' in the Figure 5 caption, and 'The mAP in the table are calculated' in the Table 1 caption.
  2. [Sec. 3.3] The sentence 'alpha_ema, alpha_ssm controls the retention ratio' should be plural ('control'), and the notation would benefit from a short explanation of how the SSM update is interleaved with the EMA update during training iterations.
  3. [Figure 2] The caption mentions a 'Spectral Student', but the text describes a single student model; please clarify whether there are two students or whether this is a labeling inconsistency.
  4. [Sec. 5] The conclusion says the method is demonstrated effective on FLIR-ADAS and FLIR, but it does not mention the absence of error bars or the grid-search selection of hyperparameters; a brief limitation statement would be appropriate.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper's claims are empirical mAP improvements, and no prediction reduces to a fitted input or to a self-citation.

full rationale

The SS-DC paper makes empirical performance claims (mAP gains on FLIR and a newly formed FLIR-ADAS protocol) that are measured against reproduced baselines; it does not derive a first-principles result whose conclusion is an input in disguise. The main candidate for a definitional worry is the self-distillation decoupling loss in Eq. (5), which defines DI/DS by internal correlation across repeated filtering rather than by any external domain-class supervision, and the hard-mode idempotence claim relies on the semi-group property of Gaussians even though a squared mask is not the same mask. Those are validity/identifiability weaknesses—the loss admits trivial identity-filter solutions and the ablations remove SAID as a whole—but they do not make the reported mAP numerically equivalent to the loss by construction. Likewise, the new FLIR-ADAS protocol and the hyperparameter search over k, lambda, filter-bank size, and SSM steps on that protocol create selection bias in the headline numbers, but the mAP values are still outer measurements rather than fitted parameters renamed as predictions. There are no load-bearing self-citations, no imported uniqueness theorems, and no known result being relabeled as a derivation. Therefore the circularity score is 0.

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

The framework rests on standard Fourier and Gaussian filter mathematics plus several domain assumptions about spectral style/content separation and pseudo-label reliability. The main ad hoc element is the claim that a soft Pearson-correlation loss enforces idempotence in the soft and free SAID modes. No new physical entities are introduced; DS tokens are architecture components rather than independently evidenced entities.

free parameters (6)
  • Number of filters N in SAID Soft mode filter bank = 100
    Swept over 10, 20, 50, 100, 200, 500; best mAP 48.02 at N=100.
  • Decoupling loss constraint strength k = 2
    Grid searched over k=1,2,3 with lambda_dcp values; best mAP at k=2.
  • Decoupling loss coefficient lambda_dcp = 50
    Swept over 5, 10, 50, 100; best mAP 48.02 at lambda_dcp=50.
  • Student Stabilization Module step size = 500 iterations
    Ablation in Fig. 5 tested 200, 500, 1000, and no SSM; 500 iterations chosen as best.
  • SSM retention ratio alpha_ssm = 0.5
    Set to 0.5 in Eq. 3 for the student stabilization mixing.
  • Adaptive fusion coefficients alpha_l = learned, unspecified
    Used in Eq. 6 and Eq. 7 for each feature level; treated as adaptive but no initialization or constraint details are given.
assumptions (5)
  • domain assumption 2D Fourier transform decomposition into amplitude and phase, with low/high frequency bands carrying style and content information respectively
    The motivation in Sec. 1 and Sec. 3.2 assumes spectral bands align with domain-invariant and domain-specific content; this is a heuristic, not a proven equivalence for object detection features.
  • standard math Gaussian filters form an idempotent semi-group
    Used in SAID-Hard mode to argue that the Gaussian high-pass filter is idempotent.
  • ad hoc to paper The Pearson-correlation decoupling loss in Eq. 5 makes soft and free SAID filters approximately idempotent and separates DI from DS in the intended way
    The paper claims idempotence is guaranteed by the self-distillation loss, but Eq. 5 is only a soft correlation objective with no convergence or separation guarantee.
  • domain assumption Mean Teacher with EMA and SSM produces reliable pseudo-labels on unlabeled infrared target data
    The teacher-student mutual learning stage assumes weakly augmented IR images yield pseudo-labels accurate enough to train the student; this is a standard self-training assumption that is not proved.
  • domain assumption The RGB source domain contains multiple subdomains whose DI/DS decomposition transfers to the IR target domain
    Central motivation in Sec. 1; the paper does not quantitatively validate that the learned decoupling transfers across the RGB-IR gap.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SS-DC: Spatial-Spectral Decoupling and Coupling Across Visible-Infrared Gap for Domain Adaptive Object Detection." pith.science (2026). https://pith.science/paper/3GDSFAZB

@misc{pith2026250712017,
  author       = {Pith},
  title        = {Pith review of: SS-DC: Spatial-Spectral Decoupling and Coupling Across Visible-Infrared Gap for Domain Adaptive Object Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3GDSFAZB}},
  note         = {Machine review of arXiv:2507.12017}
}
read the original abstract

Unsupervised domain adaptive object detection (UDAOD) from the visible domain to the infrared (RGB-IR) domain is challenging. Existing methods regard the RGB domain as a unified domain and neglect the multiple subdomains within it, such as daytime, nighttime, and foggy scenes. We argue that decoupling the domain-invariant (DI) and domain-specific (DS) features across these multiple subdomains is beneficial for RGB-IR domain adaptation. To this end, this paper proposes a new SS-DC framework based on a decoupling-coupling strategy. In terms of decoupling, we design a Spectral Adaptive Idempotent Decoupling (SAID) module in the aspect of spectral decomposition. Due to the style and content information being highly embedded in different frequency bands, this module can decouple DI and DS components more accurately and interpretably. A novel filter bank-based spectral processing paradigm and a self-distillation-driven decoupling loss are proposed to improve the spectral domain decoupling. In terms of coupling, a new spatial-spectral coupling method is proposed, which realizes joint coupling through spatial and spectral DI feature pyramids. Meanwhile, this paper introduces DS from decoupling to reduce the domain bias. Extensive experiments demonstrate that our method can significantly improve the baseline performance and outperform existing UDAOD methods on multiple RGB-IR datasets, including a new experimental protocol proposed in this paper based on the FLIR-ADAS dataset.

Figures

Figures reproduced from arXiv: 2507.12017 by the authors.

Figure 1
Figure 1. (a) and (b) show the impact of UDAOD methods for [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The SS-DC framework is based on the Mean Teacher framework. The Spectral Student uses the SAID module for self-distillation [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The SAID design includes: Hard mode uses a Gaussian low-pass filter (GLPF) and a Gaussian high-pass filter (GHPF). Free mode contains a lightweight U-Net. Soft mode is based on a Filter Bank design, combining the strengths of the first two modes. We feed the amplitude and phase spectrum into a con￾volutional flow to get a spectrum feature, while maintain￾ing controllable parameters. The feature regresses a weight ve… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Visualization comparison of model inference results for [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Ablation studies on RGB→IR FLIR-ADAS for different Student Stabilisation Module Step [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

66 extracted references · 54 canonical work pages

  1. [1]

    Human detection in aerial thermal images using faster r-cnn and ssd algorithms

    KR Akshatha, A Kotegar Karunakar, Satish B Shenoy, Ab- hilash K Pai, Nikhil Hunjanal Nagaraj, and Sambhav Singh Rohatgi. Human detection in aerial thermal images using faster r-cnn and ssd algorithms. Electronics, 11(7):1151,

    work page

  2. [2]

    Famnet: Frequency-aware matching network for cross- domain few-shot medical image segmentation

    Yuntian Bo, Yazhou Zhu, Lunbo Li, and Haofeng Zhang. Famnet: Frequency-aware matching network for cross- domain few-shot medical image segmentation. ArXiv, abs/2412.09319, 2024. 2

    arXiv 2024

  3. [3]

    Heisen- berg’s uncertainty principle

    Paul Busch, Teiko Heinonen, and Pekka Lahti. Heisen- berg’s uncertainty principle. Physics reports, 452(6):155– 176, 2007. 4

    work page 2007

  4. [4]

    Exploring object relation in mean teacher for cross-domain detection

    Qi Cai, Yingwei Pan, Chong-Wah Ngo, Xinmei Tian, Lingyu Duan, and Ting Yao. Exploring object relation in mean teacher for cross-domain detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 11457–11466, 2019. 2

    work page 2019

  5. [5]

    Contrastive mean teacher for domain adaptive ob- ject detectors

    Shengcao Cao, Dhiraj Joshi, Liang-Yan Gui, and Yu-Xiong Wang. Contrastive mean teacher for domain adaptive ob- ject detectors. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 23839– 23848, 2023. 1, 2

    work page 2023

  6. [6]

    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. 2, 3

    2020

  7. [7]

    Label matching semi-supervised object detection

    Binbin Chen, Weijie Chen, Shicai Yang, Yunyi Xuan, Jie Song, Di Xie, Shiliang Pu, Mingli Song, and Yueting Zhuang. Label matching semi-supervised object detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 14381–14390, 2022. 2

    work page 2022

  8. [8]

    Cross Domain Object Detection via Multi-Granularity Confidence Alignment based Mean Teacher

    Jiangming Chen, Li Liu, Wanxia Deng, Zhen Liu, Yu Liu, Yingmei Wei, and Yongxiang Liu. Cross domain object detection via multi-granularity confidence alignment based mean teacher. arXiv preprint arXiv:2407.07780, 2024. 2

    work page Pith review arXiv 2024

Show all 66 references
  1. [9]

    Datr: Unsu- pervised domain adaptive detection transformer with dataset- level adaptation and prototypical alignment

    Liang Chen, Jianhong Han, and Yupei Wang. Datr: Unsu- pervised domain adaptive detection transformer with dataset- level adaptation and prototypical alignment. IEEE Transac- tions on Image Processing, 2025. 1, 2, 6, 7

    2025

  2. [10]

    Neural feature search for rgb-infrared per- son re-identification

    Yehansen Chen, Lin Wan, Zhihang Li, Qianyan Jing, and Zongyuan Sun. Neural feature search for rgb-infrared per- son re-identification. In Proceedings of the IEEE/CVF con- ference on computer vision and pattern recognition , pages 587–597, 2021. 1

    2021

  3. [11]

    Mixed pseudo labels for semi-supervised ob- ject detection

    Zeming Chen, Wenwei Zhang, Xinjiang Wang, Kai Chen, and Zhi Wang. Mixed pseudo labels for semi-supervised ob- ject detection. arXiv preprint arXiv:2312.07006, 2023. 7

    2023 arXiv

  4. [12]

    Every pixel matters: Center-aware feature alignment for domain adaptive object detector

    Hsu Cheng Chun, Tsai Yi-Hsuan, Lin Yen-Yu, and Yang Ming-Hsuan. Every pixel matters: Center-aware feature alignment for domain adaptive object detector. In European Conference on Computer Vision, 2020. 7

    2020

  5. [13]

    Xception: Deep learning with depthwise separable convolutions

    Franc ¸ois Chollet. Xception: Deep learning with depthwise separable convolutions. In Proceedings of the IEEE con- ference on computer vision and pattern recognition , pages 1251–1258, 2017. 5

    2017

  6. [14]

    Denodet: Attention as deformable multi-subspace feature denoising for target detection in SAR images

    Yimian Dai, Minrui Zou, Yuxuan Li, Xiang Li, Kang Ni, and Jian Yang. Denodet: Attention as deformable multi-subspace feature denoising for target detection in SAR images. CoRR, abs/2406.02833, 2024. 2

    2024 arXiv

  7. [15]

    Saquib Sarfraz, and Mohsen Ali

    Muhammad Sohail Danish, Muhammad Haris Khan, Muhammad Akhtar Munir, M. Saquib Sarfraz, and Mohsen Ali. Improving single domain-generalized object detection: A focus on diversification and alignment. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog...

    2024

  8. [16]

    Imagenet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition, pages 248–255. Ieee, 2009. 6

    2009

  9. [17]

    Un- biased mean teacher for cross-domain object detection

    Jinhong Deng, Wen Li, Yuhua Chen, and Lixin Duan. Un- biased mean teacher for cross-domain object detection. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 4091–4101, 2021. 2

    2021

  10. [18]

    Harmo- nious teacher for cross-domain object detection

    Jinhong Deng, Dongli Xu, Wen Li, and Lixin Duan. Harmo- nious teacher for cross-domain object detection. In Proceed- ings of the IEEE/CVF conference on computer vision and pattern recognition, pages 23829–23838, 2023. 1, 2, 6, 7

    2023

  11. [19]

    D3t: Distinctive dual-domain teacher zigzagging across rgb- thermal gap for domain-adaptive object detection

    Dinh Phat Do, Taehoon Kim, Jaemin Na, Jiwon Kim, Keonho Lee, Kyunghwan Cho, and Wonjun Hwang. D3t: Distinctive dual-domain teacher zigzagging across rgb- thermal gap for domain-adaptive object detection. In Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Patter...

    2024

  12. [20]

    Robust mean teacher for continual and gradual test-time adaptation

    Mario D ¨obler, Robert A Marsden, and Bin Yang. Robust mean teacher for continual and gradual test-time adaptation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 7704–7714, 2023. 2

    2023

  13. [21]

    The pascal visual object classes challenge: A retrospective

    Mark Everingham, SM Ali Eslami, Luc Van Gool, Christo- pher KI Williams, John Winn, and Andrew Zisserman. The pascal visual object classes challenge: A retrospective. In- ternational journal of computer vision , 111:98–136, 2015. 1

    2015

  14. [22]

    Object detection in thermal spectrum for advanced driver-assistance systems (adas)

    Muhammad Ali Farooq, Peter Corcoran, Cosmin Rotariu, and Waseem Shariff. Object detection in thermal spectrum for advanced driver-assistance systems (adas). IEEE Access, 9:156465–156481, 2021. 2, 6

    2021

  15. [23]

    Dsd-da: Distillation-based source debiasing for domain adaptive ob- ject detection

    Yongchao Feng, Shiwei Li, Yingjie Gao, Ziyue Huang, Yanan Zhang, Qingjie Liu, and Yunhong Wang. Dsd-da: Distillation-based source debiasing for domain adaptive ob- ject detection. arXiv preprint arXiv:2311.10437 , 2023. 1, 2

    2023 arXiv

  16. [24]

    Flir thermal dataset for algorithm training

    FLIR. Flir thermal dataset for algorithm training. https: / / www . flir . in / oem / adas / adas - dataset - form/. 2018. 6

    2018

  17. [25]

    Ir reasoner: Real-time infrared object de- tection by visual reasoning

    Meryem Mine G ¨undo˘gan, Tolga Aksoy, Alptekin Temizel, and Ugur Halici. Ir reasoner: Real-time infrared object de- tection by visual reasoning. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 422–430, 2023. 1 9

    2023

  18. [26]

    Hahs-Vaughn

    Debbie L. Hahs-Vaughn. Foundational methods: descrip- tive statistics: bivariate and multivariate data (correlations, associations). In International Encyclopedia of Education (Fourth Edition), pages 734–750. Elsevier, Oxford, fourth edition edition, 2023. 5

    2023

  19. [27]

    Remote sensing teacher: Cross-domain detec- tion transformer with learnable frequency-enhanced feature alignment in remote sensing imagery

    Jianhong Han, Wenjie Yang, Yupei Wang, Liang Chen, and Zhaoyi Luo. Remote sensing teacher: Cross-domain detec- tion transformer with learnable frequency-enhanced feature alignment in remote sensing imagery. IEEE Transactions on Geoscience and Remote Sensing, 2024. 2, 7

    2024

  20. [28]

    Simplifying source- free domain adaptation for object detection: Effective self- training strategies and performance insights

    Yan Hao, Florent Forest, and Olga Fink. Simplifying source- free domain adaptation for object detection: Effective self- training strategies and performance insights. In European Conference on Computer Vision, 2024. 1

    2024

  21. [29]

    Dif- fusion domain teacher: Diffusion guided domain adaptive object detector

    Boyong He, Yuxiang Ji, Zhuoyue Tan, and Liaoni Wu. Dif- fusion domain teacher: Diffusion guided domain adaptive object detector. In Proceedings of the 32nd ACM Interna- tional Conference on Multimedia , pages 3284–3293, 2024. 1, 2

    2024

  22. [30]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceed- ings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016. 6

    2016

  23. [31]

    Bidirectional alignment for do- main adaptive detection with transformers

    Liqiang He, Wei Wang, Albert Chen, Min Sun, Cheng-Hao Kuo, and Sinisa Todorovic. Bidirectional alignment for do- main adaptive detection with transformers. In Proceedings of the IEEE/CVF International Conference on Computer Vi- sion, pages 18775–18785, 2023. 2

    2023

  24. [32]

    Ccnet: Criss-cross attention for semantic segmentation

    Zilong Huang, Xinggang Wang, Lichao Huang, Chang Huang, Yunchao Wei, and Wenyu Liu. Ccnet: Criss-cross attention for semantic segmentation. In Proceedings of the IEEE/CVF international conference on computer vision, pages 603–612, 2019. 6

    2019

  25. [33]

    The exponentially weighted moving aver- age

    J Stuart Hunter. The exponentially weighted moving aver- age. Journal of quality technology, 18(4):203–210, 1986. 3, 6

    1986

  26. [34]

    Dynamic retraining-updating mean teacher for source-free object de- tection

    Trinh Le Ba Khanh, Huy-Hung Nguyen, Long Hoang Pham, Duong Nguyen-Ngoc Tran, and Jae Wook Jeon. Dynamic retraining-updating mean teacher for source-free object de- tection. In European Conference on Computer Vision, pages 328–344. Springer, 2024. 1

    2024

  27. [35]

    Openimages: A public dataset for large-scale multi-label and multi-class im- age classification

    Ivan Krasin, Tom Duerig, Neil Alldrin, Vittorio Ferrari, Sami Abu-El-Haija, Alina Kuznetsova, Hassan Rom, Jasper Ui- jlings, Stefan Popov, Andreas Veit, et al. Openimages: A public dataset for large-scale multi-label and multi-class im- age classification. Dataset available fr...

    2017

  28. [36]

    Sampling, data transmission, and the nyquist rate

    HJ Landau. Sampling, data transmission, and the nyquist rate. Proceedings of the IEEE, 55(10):1701–1706, 1967. 4

    1967

  29. [37]

    Dn-detr: Accelerate detr training by intro- ducing query denoising

    Feng Li, Hao Zhang, Shilong Liu, Jian Guo, Lionel M Ni, and Lei Zhang. Dn-detr: Accelerate detr training by intro- ducing query denoising. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 13619–13627, 2022. 2

    2022

  30. [38]

    Improving cross-domain detection with self- supervised learning

    Kai Li, Curtis Wigington, Chris Tensmeyer, Vlad I Morariu, Handong Zhao, Varun Manjunatha, Nikolaos Barmpalios, and Yun Fu. Improving cross-domain detection with self- supervised learning. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, ...

    2023

  31. [39]

    Sigma: Semantic- complete graph matching for domain adaptive object de- tection

    Wuyang Li, Xinyu Liu, and Yixuan Yuan. Sigma: Semantic- complete graph matching for domain adaptive object de- tection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5291–5300,

  32. [40]

    Cross-domain adaptive teacher for object detection

    Yu-Jhe Li, Xiaoliang Dai, Chih-Yao Ma, Yen-Cheng Liu, Kan Chen, Bichen Wu, Zijian He, Kris Kitani, and Peter Vajda. Cross-domain adaptive teacher for object detection. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 7581–7590, 2022. 1, 2

    2022

  33. [41]

    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 Computer vision–ECCV 2014: 13th European conference, zurich, Switzerland, September 6-12, 2014, proceeding...

    2014

  34. [42]

    Periodically exchange teacher-student for source-free object detection

    Qipeng Liu, Luojun Lin, Zhifeng Shen, and Zhifeng Yang. Periodically exchange teacher-student for source-free object detection. In Proceedings of the IEEE/CVF international conference on computer vision, pages 6414–6424, 2023. 1

    2023

  35. [43]

    Replicable uniformity test- ing

    Sihan Liu and Christopher Ye. Replicable uniformity test- ing. Advances in Neural Information Processing Systems , 37:32039–32075, 2024. 6

    2024

  36. [44]

    Unbiased teacher for semi-supervised object detec- tion

    Yen-Cheng Liu, Chih-Yao Ma, Zijian He, Chia-Wen Kuo, Kan Chen, Peizhao Zhang, Bichen Wu, Zsolt Kira, and Peter Vajda. Unbiased teacher for semi-supervised object detec- tion. arXiv preprint arXiv:2102.09480, 2021. 2, 3

    2021 arXiv

  37. [45]

    Bridging the invisible and visible world: Translation between rgb and ir images through con- tour cycle gan

    Yawen Lu and Guoyu Lu. Bridging the invisible and visible world: Translation between rgb and ir images through con- tour cycle gan. In 2021 17th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS) , pages 1–8. IEEE, 2021. 1

    2021

  38. [46]

    Few-shot adaptive object detection with cross-domain cutmix

    Yuzuru Nakamura, Yasunori Ishii, Yuki Maruyama, and Takayoshi Yamashita. Few-shot adaptive object detection with cross-domain cutmix. In Proceedings of the Asian Con- ference on Computer Vision, pages 1350–1367, 2022. 3

    2022

  39. [47]

    The fast Fourier transform

    Henri J Nussbaumer and Henri J Nussbaumer. The fast Fourier transform. Springer, 1982. 3

    1982

  40. [48]

    U- net: Convolutional networks for biomedical image segmen- tation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U- net: Convolutional networks for biomedical image segmen- tation. In Medical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, October 5-9, 2015, proceedings, pa...

    2015

  41. [49]

    Objects365: A large-scale, high-quality dataset for object detection

    Shuai Shao, Zeming Li, Tianyuan Zhang, Chao Peng, Gang Yu, Xiangyu Zhang, Jing Li, and Jian Sun. Objects365: A large-scale, high-quality dataset for object detection. In Pro- ceedings of the IEEE/CVF international conference on com- puter vision, pages 8430–8439, 2019. 1

    2019

  42. [50]

    Assaf Shocher, Amil Dravid, Yossi Gandelsman, Inbar Mosseri, Michael Rubinstein, and Alexei A. Efros. Idem- potent generative network, 2023. 4

    2023

  43. [51]

    Mean teachers are better role models: Weight-averaged consistency targets improve 10 semi-supervised deep learning results

    Antti Tarvainen and Harri Valpola. Mean teachers are better role models: Weight-averaged consistency targets improve 10 semi-supervised deep learning results. Advances in neural information processing systems, 30, 2017. 2

    2017

  44. [52]

    Source-free domain adaptation for yolo object detection

    Simon Varailhon, Masih Aminbeidokhti, Marco Pedersoli, and Eric Granger. Source-free domain adaptation for yolo object detection. arXiv preprint arXiv:2409.16538, 2024. 3, 6, 8

    2024 arXiv

  45. [53]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszko- reit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. Advances in neural information processing systems, 30, 2017. 2

    2017

  46. [54]

    Meta-uda: Unsupervised domain adaptive thermal object detection using meta-learning

    Vibashan Vs, Domenick Poster, Suya You, Shuowen Hu, and Vishal M Patel. Meta-uda: Unsupervised domain adaptive thermal object detection using meta-learning. In proceed- ings of the IEEE/CVF winter conference on applications of computer vision, pages 1412–1423, 2022. 3

    2022

  47. [55]

    Eca-net: Efficient channel at- tention for deep convolutional neural networks

    Qilong Wang, Banggu Wu, Pengfei Zhu, Peihua Li, Wang- meng Zuo, and Qinghua Hu. Eca-net: Efficient channel at- tention for deep convolutional neural networks. In Proceed- ings of the IEEE/CVF conference on computer vision and pattern recognition, pages 11534–11542, 2020. 5

    2020

  48. [56]

    Ef- ficient teacher: Semi-supervised object detection for yolov5

    Bowen Xu, Mingtao Chen, Wenlong Guan, and Lulu Hu. Ef- ficient teacher: Semi-supervised object detection for yolov5. arXiv preprint arXiv:2302.07577, 2023. 7

    2023 arXiv

  49. [57]

    End-to- end semi-supervised object detection with soft teacher

    Mengde Xu, Zheng Zhang, Han Hu, Jianfeng Wang, Lijuan Wang, Fangyun Wei, Xiang Bai, and Zicheng Liu. End-to- end semi-supervised object detection with soft teacher. In Proceedings of the IEEE/CVF international conference on computer vision, pages 3060–3069, 2021. 2

    2021

  50. [58]

    Idempotence and percep- tual image compression

    Tongda Xu, Ziran Zhu, Dailan He, Yanghao Li, Lina Guo, Yuanyuan Wang, Zhe Wang, Hongwei Qin, Yan Wang, Jingjing Liu, and Ya-Qin Zhang. Idempotence and percep- tual image compression. In The Twelfth International Con- ference on Learning Representations, 2024. 4

    2024

  51. [59]

    H2fa r-cnn: Holistic and hierarchical feature align- ment for cross-domain weakly supervised object detection

    Yunqiu Xu, Yifan Sun, Zongxin Yang, Jiaxu Miao, and Yi Yang. H2fa r-cnn: Holistic and hierarchical feature align- ment for cross-domain weakly supervised object detection. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 14329–14339...

    2022

  52. [60]

    Rethinking weak- to-strong augmentation in source-free domain adaptive ob- ject detection

    Jiuzheng Yang, Song Tang, Yangkuiyi Zhang, Shuaifeng Li, Mao Ye, Jianwei Zhang, and Xiatian Zhu. Rethinking weak- to-strong augmentation in source-free domain adaptive ob- ject detection. arXiv preprint arXiv:2410.05557, 2024. 1

  53. [61]

    Chan- nel augmentation for visible-infrared re-identification

    Mang Ye, Zesen Wu, Cuiqun Chen, and Bo Du. Chan- nel augmentation for visible-infrared re-identification. IEEE Transactions on Pattern Analysis and Machine Intelligence, 46(4):2299–2315, 2023. 1

    2023

  54. [62]

    Learning spectral- decomposited tokens for domain generalized semantic seg- mentation

    Jingjun Yi, Qi Bi, Hao Zheng, Haolan Zhan, Wei Ji, Yawen Huang, Yuexiang Li, and Yefeng Zheng. Learning spectral- decomposited tokens for domain generalized semantic seg- mentation. In Proceedings of the 32nd ACM International Conference on Multimedia, pages 8159–8168, 2024. 2

    2024

  55. [63]

    Dino: Detr with improved denoising anchor boxes for end-to-end object detection

    Hao Zhang, Feng Li, Shilong Liu, Lei Zhang, Hang Su, Jun Zhu, Lionel M Ni, and Heung-Yeung Shum. Dino: Detr with improved denoising anchor boxes for end-to-end object detection. arXiv preprint arXiv:2203.03605, 2022. 2, 7

    2022 arXiv

  56. [64]

    Semi-detr: Semi-supervised object detection with de- tection transformers

    Jiacheng Zhang, Xiangru Lin, Wei Zhang, Kuo Wang, Xiao Tan, Junyu Han, Errui Ding, Jingdong Wang, and Guanbin Li. Semi-detr: Semi-supervised object detection with de- tection transformers. In Proceedings of the IEEE/CVF con- ference on computer vision and pattern recognition ,...

    2023

  57. [65]

    Masked retraining teacher- student framework for domain adaptive object detection

    Zijing Zhao, Sitong Wei, Qingchao Chen, Dehui Li, Yifan Yang, Yuxin Peng, and Yang Liu. Masked retraining teacher- student framework for domain adaptive object detection. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 19039–19049, 2023. 1

    2023

  58. [66]

    Meta- exploiting frequency prior for cross-domain few-shot learn- ing

    Fei Zhou, Peng Wang, Lei Zhang, Zhenghua Chen, Wei Wei, Chen Ding, Guosheng Lin, and Yanning Zhang. Meta- exploiting frequency prior for cross-domain few-shot learn- ing. In Advances in Neural Information Processing Systems, pages 116783–116814. Curran Associates, Inc., 2024. 2 11

    2024

Pith tools

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