Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

Spectral-Aware Global Fusion for RGB-Thermal Semantic Segmentation

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The paper claims that SGFNet, a spectral-aware fusion network, outperforms prior methods on RGB-thermal semantic segmentation benchmarks by concentrating cross-modal fusion on high-frequency features.

desk verdict Solid incremental spectral-fusion paper with a plausible ablation, but the SOTA claim rests on unmatched baselines and omits the authors' own Sigma; needs matched retraining and variance reporting. read the letter →

arxiv 2505.15491 v1 pith:RZ5CT5IS submitted 2025-05-21 cs.CV

Ce Zhang , Zifu Wan , Simon Stepputtis , Katia Sycara , Yaqi Xie This is my paper · ORCID
classification cs.CV
keywords RGB-Tsemanticsegmentationspectral-awarefusiondiscretecosinetransformcross-modalattentionthermalimagingmulti-modalautonomousdriving
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 claims that RGB and thermal features fuse better when the fusion is guided by frequency content rather than by raw appearance. The proposed network, SGFNet, treats the two modalities as sharing broad low-frequency context while each modality carries unique high-frequency edges and textures, and it explicitly couples those high-frequency components across modalities. Through a discrete cosine transform channel-weighting step, a channel attention over the joint 2C features, and a global cross-modal spatial attention, SGFNet reports the best mean accuracy and mean IoU on both the MFNet and PST900 benchmarks. If correct, this gives a practical recipe for making semantic segmentation more reliable in low light and occlusion, where RGB alone fails.

What carries the argument

The load-bearing object is the Spectral-aware Global Fusion (SGF) module, a three-part fusion block inserted between dual ResNet-152 encoders and a cascaded BBS-Net decoder. First, features are split into N channel groups, and each group is weighted by a scalar derived from projecting the group onto a 2D discrete cosine transform basis, with the lowest frequency pair reducing to global average pooling; this produces a multi-spectral vector per channel that drives a channel-wise activation. Second, the concatenated 2C-channel enhanced features pass through a large-kernel attention (5x5 depthwise convolution, 7x7 dilated depthwise convolution, then 1x1 convolution) followed by another DCT-based channel activation, so the model reweights which channels across both modalities matter most. Third, a global cross-modal spatial attention computes interactions between all pixel pairs of the RGB and thermal streams using query, key, and value projections from each modality, with residual connections to fuse the streams. The fused map is decoded by the cascaded decoder, and auxiliary losses supervise the four early fusion stages.

What would settle it

Re-run the listed baseline methods inside SGFNet's exact training pipeline (100 epochs, batch size 2, initial learning rate 0.02, exponential decay 0.95, random flip and crop, Dice plus SoftCrossEntropy loss) on MFNet and PST900, and compare mIoU; if the margins over the strongest baselines shrink below run-to-run variance or reverse, the central outperformance claim fails.

Watch

Extended reading notes

Core claim

The central discovery the paper advances is that RGB-thermal fusion succeeds when the network explicitly couples high-frequency, modality-specific components while letting the shared low-frequency context pass through more directly. SGFNet encodes this by representing each feature map's channels as multi-spectral vectors computed with 2D DCT bases, jointly reweighting both modalities, and then applying a joint channel attention over all 2C channels and a global pixel-pair cross-attention between the two streams. Its reported results are 76.2% mAcc and 60.1% mIoU on MFNet, and 91.90% mAcc and 85.37% mIoU on PST900, the best among the methods listed in its tables. The ablation attributes the largest single gain to the spectral-aware enhancement and channel attention, with the complete system adding 5.2 mIoU over the no-component baseline on MFNet.

Load-bearing premise

The comparison assumes that the baseline numbers in Tables 1 and 2 were produced under conditions equivalent enough to SGFNet's training protocol (100 epochs, batch size 2, initial learning rate 0.02, exponential decay 0.95, random flip and crop, ImageNet-pretrained ResNet-152, 7x7 DCT basis), because the paper reruns no baseline itself and reports no variance.

Editorial extensions

If this is right

  • SGFNet reports state-of-the-art mAcc/mIoU on MFNet (76.2%/60.1%) and PST900 (91.90%/85.37%), beating the best listed prior methods by roughly 1.2 and 1.0 mIoU respectively.
  • The gains are visible on challenging classes such as Curve, Bump, Car Stop, and Guardrail on MFNet, suggesting that explicit high-frequency interaction helps boundary-like and small objects.
  • SGFNet also beats the transformer-based CMX and CMNeXt methods, which use SegFormer backbones, in MFNet mIoU while using only a ResNet-152 backbone, indicating the fusion module contributes beyond backbone capacity.
  • The efficiency comparison shows SGFNet needs 249.25 GFLOPs and 163.99M parameters, placing it between RTFNet and EAEFNet in cost while giving higher accuracy than EAEFNet on MFNet.
  • The ablation shows that removing all four components drops mIoU from 60.1% to 54.9%, implying the full combination, not any single module, produces the reported margin.

Reading between the lines

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

  • The spectral split suggests a transferable design principle: for any paired sensors whose high-frequency statistics differ, forcing the network to align high-frequency residuals may improve fusion; the paper does not test this on other modality pairs such as depth and RGB.
  • Because the paper reports no variance over random seeds and does not rerun any baseline in its own pipeline, the 1-2 mIoU margins may be within run-to-run noise; a direct extension is to train with multiple seeds and report error bars.
  • The 7x7 DCT basis and the number of channel groups N are chosen empirically; sweeping these hyperparameters would reveal whether the spectral grouping itself, rather than the added parameters and attention, carries the improvement.
  • The model's computational cost (249 GFLOPs, 164M parameters) is at the high end for embedded driving use, so an obvious undeveloped direction is distilling or pruning the spectral and cross-attention modules for real-time deployment.
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

3 major / 4 minor

Summary. The paper proposes SGFNet, an RGB-thermal semantic segmentation network built around a spectral-aware fusion module. The module includes a DCT-based channel enhancement that weights RGB and thermal features by spectral components, a spectral-aware channel attention mechanism applied across the concatenated multi-modal features, a global cross-modal spatial attention operation, and deep supervision on a BBS-Net-style decoder. The authors evaluate SGFNet on the MFNet and PST900 datasets and report state-of-the-art results (60.1 mIoU on MFNet and 85.37 mIoU on PST900), together with ablations and efficiency comparisons.

Significance. If the reported results are reliable, SGFNet provides a useful, incremental contribution to RGB-T semantic segmentation: the idea of explicitly encouraging interaction of high-frequency, modality-specific components is motivated by qualitative examples and the ablations show consistent gains from the proposed modules. The paper also provides transparent efficiency numbers and a clean architectural description. However, the central claim is an empirical benchmark superiority that is currently supported only by an uncontrolled comparison against transcribed baseline numbers, with no variance estimates, and the authors' own recent Sigma method is omitted from the tables. The contribution is therefore plausible but not yet established at the level claimed.

major comments (3)
  1. [Tables 1 and 2; Sec. 3.1] The central claim that SGFNet outperforms state-of-the-art methods rests on a comparison in which all baseline numbers are copied from prior publications with their own training protocols, while SGFNet is trained under the protocol in Sec. 3.1 (100 epochs, batch size 2, initial learning rate 0.02, exponential decay 0.95, random flip/crop, ImageNet-pretrained ResNet-152). Since the reported margins are only +1.2 mIoU on MFNet and +0.97 mIoU on PST900, and no baseline is retrained under SGFNet's protocol, these margins could easily be within run-to-run variation caused by different schedules, augmentations, or loss terms. Please provide matched-condition comparisons, for example by retraining at least the strongest baselines under the same protocol, or by reporting multiple seeds with means and variances for both SGFNet and the baselines.
  2. [Tables 1 and 2; reference [3]] The authors' own Sigma network (WACV 2025, reference [3]) is cited in the Introduction but is absent from both comparison tables. Because Sigma is a recent RGB-T semantic segmentation method and the paper claims to outperform state-of-the-art methods, omitting it from the benchmark comparisons is a substantive gap: add Sigma to Tables 1 and 2 or explicitly justify its exclusion. This omission is especially important if Sigma's published numbers exceed SGFNet's on either dataset.
  3. [Sec. 2.1, Eqs. (2)-(3)] The method description omits implementation details needed to reproduce the spectral grouping: the number of groups N and the assignment of frequency pairs (f^i_h, f^i_w) to each group are not specified, and it is unclear whether the SGF module is applied at every encoder scale or only one scale (Figure 2 says multi-scale, while Sec. 2.1 says 'at a particular scale'). Please specify these choices; without N and the frequency assignment, the DCT-based channel score in Eq. (3) is not fully defined.
minor comments (4)
  1. [Sec. 2.1, Eq. (5)] The expression Sigmoid(1-Q) is not the complement of Sigmoid(Q); if the intent is to retain features that are suppressed by the enhancement, the complementary gate should be written as 1 - Sigmoid(Q), or the paper should clarify why the shifted sigmoid is intentional.
  2. [Sec. 2.3, Eq. (13)] Eq. (13) defines F_fuse_T = F_att_T, but Sec. 2.4 only describes using F_fuse_RGB for the preliminary prediction; if F_fuse_T is unused by the decoder, this should be stated explicitly.
  3. [Before Eq. (4)] The factor C in Q = C * Q_RGB * Q_T is unexplained; if C is the number of channels, this is a dimension-dependent scaling that should be justified or absorbed into the MLP weights.
  4. [Table 1] The column header of Table 1 is garbled ('mAcc mIoUAcc IoU ...'); please split the accuracy and IoU columns cleanly for each class and for the aggregate metrics.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation; the SOTA claim is an empirical benchmark comparison with only a contextual self-citation.

full rationale

No load-bearing step of SGFNet reduces by construction to its own inputs. The core fusion chain (Eqs. 1–13) builds on external, verifiable components: DCT bases from FcaNet [11], large-kernel attention [12], transformer-style cross-attention [13], and the BBS-Net decoder [14]; the paper defines its DCT and attention operations explicitly, so the spectral enhancement and fusion scores are actual network operations rather than renamed data fits. The premise that high-frequency features carry modality-specific details is an observational motivation (Fig. 1), not a conclusion derived from the benchmark numbers. The central claim that SGFNet outperforms state-of-the-art methods is an empirical comparison: Tables 1 and 2 report measured mAcc/mIoU values, and Table 3 ablates the full model against its own stripped version, which is a controlled internal comparison. The only author self-citation is reference [3] (Sigma), cited in the introduction as general related work; it is neither used in the method nor included in the comparison tables, so it is not load-bearing. Whether the baseline numbers were produced under training protocols matched to SGFNet is a validity/correctness concern about uncontrolled comparisons, not a circularity of the derivation chain; the prediction is not equivalent to the input by construction. Since no specific equation-to-equation reduction or fitted-input-renamed-as-prediction can be exhibited, the circularity score is 0.

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

The numerical claims rest on standard deep-learning assumptions: ImageNet-pretrained features transfer to thermal images, DCT coefficients summarize channel importance, the fixed BBS-Net decoder is adequate, and the hand-set training configuration is appropriate. The method also depends on an unexplained product rule for joint channel scores (Eq. 4) and on an empirical choice of DCT basis size (7,7). No new physical entities are introduced.

free parameters (6)
  • DCT basis size (H', W') = (7, 7)
    Set empirically in Section 3.1; controls which DCT frequencies are pooled and affects the spectral channel attention of Eq. (2)-(3).
  • Number of spectral groups N and frequency-pair assignment = unspecified
    Eq. (2) partitions channels into N groups, each assigned a unique DCT frequency pair; N and the assignment rule are never stated, so the central spectral module is under-specified.
  • Cross-attention query/key channel reduction C' = C/8 = C/8
    Chosen to reduce computational cost (Section 2.3); no ablation is provided for this ratio.
  • Training hyperparameters (epochs, batch, learning rate, decay) = 100 epochs, batch size 2, lr 0.02, decay 0.95
    Section 3.1; not justified by theory and may affect comparison to baselines.
  • Encoder choice ResNet-152 = ResNet-152 ImageNet-pretrained
    Table 5 shows performance scales with backbone; comparisons to methods using SegFormer or MobileNet are confounded by backbone.
  • Loss weighting = equal weights (1.0 for each term)
    Section 2.4 sums Dice and SoftCrossEntropy losses without a weighting scheme or sensitivity analysis.
assumptions (6)
  • standard math 2D DCT coefficient vectors are a sufficient statistic for channel-wise importance in feature maps (from FcaNet [11])
    Eq. (1)-(3) treat DCT components as the representation used to derive channel activation scores; no justification specific to RGB-T is given.
  • domain assumption High-frequency components capture modality-specific edges and textures while low-frequency components capture shared context
    Section 1 and Fig. 1 present this as an observation based on two Grad-CAM visualizations, not a measured property used to set constants.
  • domain assumption ImageNet-pretrained ResNet-152 features transfer to thermal images
    Section 3.1 initializes both RGB and thermal encoders with ImageNet weights; thermal images are outside the ImageNet distribution.
  • domain assumption BBS-Net cascaded decoder [14] is a suitable fixed decoder for RGB-T fusion
    Section 2.4 adopts BBS-Net without ablation or RGB-T-specific adaptation.
  • ad hoc to paper Equal weighting of Dice and SoftCrossEntropy losses and of intermediate deep-supervision terms is appropriate
    Section 2.4 defines total loss as a sum of Eq. (14) and (15) with no weighting scheme or sensitivity analysis.
  • ad hoc to paper Multiplying per-modality DCT scores as Q = C * Q_RGB * Q_T yields a meaningful joint channel score
    Eq. (4) introduces a product rule and a channel-count scaling C without derivation or ablation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Spectral-Aware Global Fusion for RGB-Thermal Semantic Segmentation." pith.science (2026). https://pith.science/paper/RZ5CT5IS

@misc{pith2026250515491,
  author       = {Pith},
  title        = {Pith review of: Spectral-Aware Global Fusion for RGB-Thermal Semantic Segmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RZ5CT5IS}},
  note         = {Machine review of arXiv:2505.15491}
}
read the original abstract

Semantic segmentation relying solely on RGB data often struggles in challenging conditions such as low illumination and obscured views, limiting its reliability in critical applications like autonomous driving. To address this, integrating additional thermal radiation data with RGB images demonstrates enhanced performance and robustness. However, how to effectively reconcile the modality discrepancies and fuse the RGB and thermal features remains a well-known challenge. In this work, we address this challenge from a novel spectral perspective. We observe that the multi-modal features can be categorized into two spectral components: low-frequency features that provide broad scene context, including color variations and smooth areas, and high-frequency features that capture modality-specific details such as edges and textures. Inspired by this, we propose the Spectral-aware Global Fusion Network (SGFNet) to effectively enhance and fuse the multi-modal features by explicitly modeling the interactions between the high-frequency, modality-specific features. Our experimental results demonstrate that SGFNet outperforms the state-of-the-art methods on the MFNet and PST900 datasets.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. DRPFNet: Dual-domain Residual Progressive Fusion Network for RGB-Thermal Object Detection

    cs.CV 2026-08 conditional novelty 4.0 of 10

    DRPFNet reports 97.8% mAP50 on LLVIP and 88.7% on M3FD with a dual-domain, progressive fusion network for RGB-T detection.

Reference graph

Works this paper leans on

33 extracted references · 32 canonical work pages · cited by 1 Pith paper

  1. [3]

    Experimental Settings Datasets and Metrics

    EXPERIMENTS 3.1. Experimental Settings Datasets and Metrics . Following the literature [8, 17], we use the MFNet [4] and PST900 [9] datasets for benchmarking differ- ent RGB-T segmentation methods. The MFNet dataset consists of 1,569 pairs of RGB and thermal images, each with a resolution of 480 × 640. We follow the splitting scheme [8, 17] to use 50% of ...

    work page

  2. [1]

    INTRODUCTION Semantic segmentation, which involves pixel-level scene under- standing, is crucial for how autonomous agents perceive and interact with their environment. It empowers autonomous driving vehi- cles to distinguish between roads, pedestrians, and obstacles in real-time, ensuring safe navigation through complex urban environ- ments [1, 2, 3]. Re...

    work page Pith review arXiv 2025

  3. [2]

    Spectral-Aware Feature Enhancement We utilize the ResNet-152 [10] encoders to first extract features from both RGB and thermal images at a particular scale

    METHOD 2.1. Spectral-Aware Feature Enhancement We utilize the ResNet-152 [10] encoders to first extract features from both RGB and thermal images at a particular scale. These features, denoted as FRGB/T = fResNet(IRGB/T), serve as the inputs for the SGF module. As shown in Figure 3, to enhance the interaction of the extracted features, we jointly assign w...

    work page

  4. [4]

    CONCLUSION In this work, we propose Spectral-aware Global Fusion Network (SGFNet) for RGB-T semantic segmentation. Specifically, we en- hance the precision of segmentation by refining the fusion process of multi-modal features, focusing on amplifying the interactions among high-frequency features that contain distinct and modality- specific details. We al...

    work page

  5. [5]

    ACKNOWLEDGMENTS This work has been funded in part by Army Research Laboratory (ARL) award W911NF-23-2-0007 and W911QX-24-F-0049, DARPA award FA8750-23-2-1015, and ONR award N00014-23-1-2840

    work page

  6. [6]

    Deep multi-modal object detection and semantic segmentation for autonomous driving: Datasets, methods, and challenges,

    Di Feng, Christian Haase-Sch ¨utz, Lars Rosenbaum, Heinz Hertlein, Claudius Glaeser, Fabian Timm, Werner Wiesbeck, and Klaus Dietmayer, “Deep multi-modal object detection and semantic segmentation for autonomous driving: Datasets, methods, and challenges,” IEEE TITS, vol. 22, no. 3, pp. 1341– 1360, 2020

    work page 2020

  7. [7]

    Rtfnet: Rgb- thermal fusion network for semantic segmentation of urban scenes,

    Yuxiang Sun, Weixun Zuo, and Ming Liu, “Rtfnet: Rgb- thermal fusion network for semantic segmentation of urban scenes,” IEEE RAL, vol. 4, no. 3, pp. 2576–2583, 2019

    work page 2019

  8. [8]

    Sigma: Siamese mamba network for multi-modal semantic segmentation,

    Zifu Wan, Pingping Zhang, Yuhao Wang, Silong Yong, Si- mon Stepputtis, Katia Sycara, and Yaqi Xie, “Sigma: Siamese mamba network for multi-modal semantic segmentation,” in WACV. IEEE, 2025, pp. 1734–1744

    work page 2025

Show all 33 references
  1. [9]

    Mfnet: Towards real-time semantic segmentation for autonomous vehicles with multi- spectral scenes,

    Qishen Ha, Kohei Watanabe, Takumi Karasawa, Yoshitaka Ushiku, and Tatsuya Harada, “Mfnet: Towards real-time semantic segmentation for autonomous vehicles with multi- spectral scenes,” in IROS. IEEE, 2017, pp. 5108–5115

    2017

  2. [10]

    Feanet: Feature-enhanced attention network for rgb-thermal real-time semantic segmentation,

    Fuqin Deng, Hua Feng, Mingjian Liang, Hongmin Wang, Yong Yang, Yuan Gao, Junfeng Chen, Junjie Hu, Xiyue Guo, and Tin Lun Lam, “Feanet: Feature-enhanced attention network for rgb-thermal real-time semantic segmentation,” in IROS. IEEE, 2021, pp. 4467–4473

    2021

  3. [11]

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

    Ramprasaath R Selvaraju, Michael Cogswell, Abhishek Das, Ramakrishna Vedantam, Devi Parikh, and Dhruv Batra, “Grad- cam: Visual explanations from deep networks via gradient- based localization,” in ICCV, 2017, pp. 618–626

    2017

  4. [12]

    Bisenet: Bilateral segmentation network for real-time semantic segmentation,

    Changqian Yu, Jingbo Wang, Chao Peng, Changxin Gao, Gang Yu, and Nong Sang, “Bisenet: Bilateral segmentation network for real-time semantic segmentation,” inECCV, 2018, pp. 325– 341

    2018

  5. [13]

    Explicit attention-enhanced fusion for rgb-thermal perception tasks,

    Mingjian Liang, Junjie Hu, Chenyu Bao, Hua Feng, Fuqin Deng, and Tin Lun Lam, “Explicit attention-enhanced fusion for rgb-thermal perception tasks,” IEEE RAL, vol. 8, no. 7, pp. 4060–4067, 2023

    2023

  6. [14]

    Pst900: Rgb- thermal calibration, dataset and segmentation network,

    Shreyas S Shivakumar, Neil Rodrigues, Alex Zhou, Ian D Miller, Vijay Kumar, and Camillo J Taylor, “Pst900: Rgb- thermal calibration, dataset and segmentation network,” in ICRA. IEEE, 2020, pp. 9441–9447

    2020

  7. [15]

    Deep residual learning for image recognition,

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun, “Deep residual learning for image recognition,” in CVPR, 2016, pp. 770–778

    2016

  8. [16]

    Fcanet: Fre- quency channel attention networks,

    Zequn Qin, Pengyi Zhang, Fei Wu, and Xi Li, “Fcanet: Fre- quency channel attention networks,” in ICCV, 2021, pp. 783– 792

    2021

  9. [17]

    Visual attention network,

    Meng-Hao Guo, Cheng-Ze Lu, Zheng-Ning Liu, Ming-Ming Cheng, and Shi-Min Hu, “Visual attention network,” Compu- tational Visual Media, vol. 9, no. 4, pp. 733–752, 2023

    2023

  10. [18]

    Attention is all you need,

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Lukasz Kaiser, and Illia Polo- sukhin, “Attention is all you need,” in NeurIPS, 2017, vol. 30, pp. 6000–6010

    2017

  11. [19]

    Bbs-net: Rgb-d salient object detection with a bifur- cated backbone strategy network,

    Deng-Ping Fan, Yingjie Zhai, Ali Borji, Jufeng Yang, and Ling Shao, “Bbs-net: Rgb-d salient object detection with a bifur- cated backbone strategy network,” in ECCV. Springer, 2020, pp. 275–292

    2020

  12. [20]

    Segformer: Simple and ef- ficient design for semantic segmentation with transformers,

    Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M Alvarez, and Ping Luo, “Segformer: Simple and ef- ficient design for semantic segmentation with transformers,” NeurIPS, vol. 34, pp. 12077–12090, 2021

    2021

  13. [21]

    Abmdrnet: Adaptive- weighted bi-directional modality difference reduction network for rgb-t semantic segmentation,

    Qiang Zhang, Shenlu Zhao, Yongjiang Luo, Dingwen Zhang, Nianchang Huang, and Jungong Han, “Abmdrnet: Adaptive- weighted bi-directional modality difference reduction network for rgb-t semantic segmentation,” in CVPR, 2021, pp. 2633– 2642

    2021

  14. [22]

    Gmnet: Graded-feature multilabel-learning network for rgb-thermal urban scene semantic segmentation,

    Wujie Zhou, Jinfu Liu, Jingsheng Lei, Lu Yu, and Jenq-Neng Hwang, “Gmnet: Graded-feature multilabel-learning network for rgb-thermal urban scene semantic segmentation,” IEEE TIP, vol. 30, pp. 7790–7802, 2021

    2021

  15. [23]

    Edge-aware guidance fusion network for rgb–thermal scene parsing,

    Wujie Zhou, Shaohua Dong, Caie Xu, and Yaguan Qian, “Edge-aware guidance fusion network for rgb–thermal scene parsing,” in AAAI, 2022, vol. 36, pp. 3571–3579

    2022

  16. [24]

    Mtanet: Multitask-aware network with hierarchical multi- modal fusion for rgb-t urban scene understanding,

    Wujie Zhou, Shaohua Dong, Jingsheng Lei, and Lu Yu, “Mtanet: Multitask-aware network with hierarchical multi- modal fusion for rgb-t urban scene understanding,” IEEE TIV, vol. 8, no. 1, pp. 48–58, 2022

    2022

  17. [25]

    Dual-space graph-based interaction network for rgb- thermal semantic segmentation in electric power scene,

    Chang Xu, Qingwu Li, Xiongbiao Jiang, Dabing Yu, and Yaqin Zhou, “Dual-space graph-based interaction network for rgb- thermal semantic segmentation in electric power scene,” IEEE TCSVT, vol. 33, no. 4, pp. 1577–1592, 2022

    2022

  18. [26]

    A feature divide-and-conquer network for rgb-t semantic segmentation,

    Shenlu Zhao and Qiang Zhang, “A feature divide-and-conquer network for rgb-t semantic segmentation,” IEEE TCSVT, vol. 33, no. 6, pp. 2892–2905, 2022

    2022

  19. [27]

    Multispectral fusion trans- former network for rgb-thermal urban scene semantic segmen- tation,

    Heng Zhou, Chunna Tian, Zhenxi Zhang, Qizheng Huo, Yongqiang Xie, and Zhongbo Li, “Multispectral fusion trans- former network for rgb-thermal urban scene semantic segmen- tation,” IEEE GRSL, vol. 19, pp. 1–5, 2022

    2022

  20. [28]

    Rgb-t semantic segmentation with location, activation, and sharpening,

    Gongyang Li, Yike Wang, Zhi Liu, Xinpeng Zhang, and Dan Zeng, “Rgb-t semantic segmentation with location, activation, and sharpening,” IEEE TCSVT, vol. 33, no. 3, pp. 1223–1235, 2023

    2023

  21. [29]

    Context-aware interac- tion network for rgb-t semantic segmentation,

    Ying Lv, Zhi Liu, and Gongyang Li, “Context-aware interac- tion network for rgb-t semantic segmentation,” IEEE TMM, vol. 26, pp. 6348–6360, 2024

    2024

  22. [30]

    Cmx: Cross-modal fusion for rgb-x semantic segmentation with transformers,

    Jiaming Zhang, Huayao Liu, Kailun Yang, Xinxin Hu, Ruiping Liu, and Rainer Stiefelhagen, “Cmx: Cross-modal fusion for rgb-x semantic segmentation with transformers,” IEEE TITS, 2023

    2023

  23. [31]

    Delivering arbitrary-modal semantic segmen- tation,

    Jiaming Zhang, Ruiping Liu, Hao Shi, Kailun Yang, Simon Reiß, Kunyu Peng, Haodong Fu, Kaiwei Wang, and Rainer Stiefelhagen, “Delivering arbitrary-modal semantic segmen- tation,” in CVPR, 2023, pp. 1136–1147

    2023

  24. [32]

    Multi-interactive feature learning and a full-time multi-modality benchmark for image fusion and segmentation,

    Jinyuan Liu, Zhu Liu, Guanyao Wu, Long Ma, Risheng Liu, Wei Zhong, Zhongxuan Luo, and Xin Fan, “Multi-interactive feature learning and a full-time multi-modality benchmark for image fusion and segmentation,” in ICCV, 2023, pp. 8115– 8124

    2023

  25. [33]

    V-net: Fully convolutional neural networks for volumetric medical image segmentation,

    Fausto Milletari, Nassir Navab, and Seyed-Ahmad Ahmadi, “V-net: Fully convolutional neural networks for volumetric medical image segmentation,” in 3DV, 2016, pp. 565–571. 6

    2016

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.