Pith. sign in

REVIEW 4 major objections 6 minor 1 cited by

TeSG: Textual Semantic Guidance for Infrared and Visible Image Fusion

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read TeSG shows that dual-level textual semantics—mask and text—guide infrared-visible fusion to better support object detection and segmentation.

desk verdict New combination of mask-level and text-level guidance for infrared-visible fusion, with solid experiments, but the load-bearing mask-generation step is underspecified and unvalidated. read the letter →

arxiv 2506.16730 v1 pith:F6OHY62F submitted 2025-06-20 cs.CV

Mingrui Zhu , Xiru Chen , Xin Wei , Nannan Wang , Xinbo Gao This is my paper · ORCID
classification cs.CV
keywords infraredandvisibleimagefusiontext-guidedvision-languagemodelsemanticmaskcross-attentionobjectdetectionsegmentation
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 proposes TeSG, a fusion method that uses automatically extracted textual descriptions of a scene to guide how infrared and visible images are combined. The claim is that steering the fusion at two levels, a mask level that marks key targets and a text level that filters visual features, produces fused images that better support downstream tasks such as object detection and semantic segmentation. Across three public benchmarks the method reports state-of-the-art or competitive scores on standard fusion metrics, and best results on detection and segmentation. This matters because text guidance is flexible, requires no manual input, and offers a route from low-level pixel fusion toward semantics-aware fusion tuned for high-level vision tasks.

What carries the argument

The central mechanism is the Semantic Information Generator (SIG) combined with two attention-based modules. SIG runs a frozen BLIP encoder on the visible image to produce a textual description, removes a keyword, then compares the noise predicted by a frozen DDIM model conditioned on the full text versus the text without the keyword; the normalized and binarized difference defines a semantic mask, and the text encoder supplies text embeddings. The Mask-Guided Cross-Attention (MGCA) module splits both modalities into foreground and background using that mask, then reconstructs each modality's features from the other's via cross-attention on each region. The Text-Driven Attentional Fusion (TDAF) module concatenates the reconstructed features, applies cross-attention with the text embeddings, and uses a sigmoid gate plus learned spatial attention to weight visible versus infrared contributions pixel-wise.

What would settle it

Computing the intersection-over-union between TeSG's semantic masks and ground-truth object or segment annotations on MSRS would settle the claim: if the masks align poorly (e.g., mIoU below 0.5) while the fusion metrics remain high, then the mask-level guidance is not the source of the reported improvement.

Watch

Extended reading notes

Core claim

The central discovery claimed is that textual semantic guidance, applied both as a mask and as a text embedding, yields fused infrared-visible images with richer information, clearer edges, and improved performance on object detection and semantic segmentation compared with nine recent fusion methods. The mask guidance is derived from a frozen diffusion model's noise-difference map that isolates the region associated with a keyword in the BLIP-generated description, while the text guidance uses the full description's embedding to modulate cross-attention and gated fusion. The paper attributes the gains to jointly applying these two levels of guidance rather than to either alone, and its ablations show metric drops when each module is removed.

Load-bearing premise

The method assumes that comparing a frozen diffusion model's noise predictions with and without a keyword cleanly isolates that object's region, and that the union of the two masks reliably separates foreground from background.

Editorial extensions

If this is right

  • Fusion no longer needs manual text input: BLIP automatically supplies the descriptions, so the pipeline runs end to end on any paired input.
  • Downstream detection and segmentation improve without retraining the fusion network for a specific task, because guidance is generic text rather than a dedicated task head.
  • The same dual-level guidance could transfer to other fusion settings—such as medical or remote sensing imagery—where text can name salient structures of interest.
  • Ablations indicate that both mask and text guidance contribute, and removing either reduces information-based metrics, so future designs should keep both levels active.

Reading between the lines

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

  • A natural next test is whether the DDIM-noise-difference masks actually align with human-annotated object boundaries; the paper does not report mask IoU, so that validation remains open.
  • Because text descriptions come from BLIP on the visible image, the guidance inherits visible-light biases; in very dark scenes the description may miss thermal-only targets, a limitation the current experiments may underweight.
  • The gated fusion formula resembles learnable per-pixel alpha blending conditioned on text; one could measure how often the spatial attention weight deviates from 0.5 to see when text substantially changes fusion decisions.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper proposes TeSG, a text-guided infrared and visible image fusion method that injects textual semantics at two levels: a mask level and a text level. A frozen BLIP model produces a caption, from which a keyword V* is removed; a frozen DDIM denoiser is then used to derive a semantic mask by thresholding the difference between noise predictions with and without V*, and the mask is combined with a similarly derived infrared mask. The mask guides a Mask-Guided Cross-Attention module, and the text embedding guides a Text-Driven Attentional Fusion module. Experiments compare TeSG with nine fusion methods on MSRS, RoadScene, and LLVIP, reporting fusion metrics and downstream object detection and semantic segmentation results, plus ablations on LLVIP.

Significance. If the mask-generation mechanism is validated, this is a meaningful contribution to text-guided image fusion: it combines a recent diffusion-based localization idea (DDIM noise difference) with a two-level semantic guidance scheme, and the paper provides a broad experimental comparison including downstream tasks. The authors are careful to use frozen BLIP and DDIM components, and the experimental protocol is extensive. However, the central mechanism that is supposed to make the method semantic rather than a generic spatial weighting is not specified in enough detail to reproduce or verify, and the ablation evidence is too weak to establish that the textual semantics are what drive the reported improvements. The paper would be substantially strengthened by a mask-localization evaluation and by ablations that control for the spatial mask itself.

major comments (4)
  1. [Sec. 3.2, Eq. (2)-(3)] The mask generation pipeline is underspecified. The paper does not state how the keyword V* is selected from the BLIP caption, what threshold or normalization procedure is used in Δ(·), how many DDIM steps or which noise schedule is used, or how the single-channel infrared image I_ir is fed into a diffusion model that was presumably trained on RGB images. Without these details, Eq. (2) and Eq. (3) are not reproducible, and the reported gains cannot be independently checked.
  2. [Sec. 3.2 / Sec. 4.4] There is no validation that the generated mask M actually localizes the semantic objects named in the caption. Since the MSRS dataset has segmentation labels, the authors could compute IoU or Dice between M and the ground-truth foreground regions. This is load-bearing because the Mask-Guided Cross-Attention module consumes M directly; if M is an arbitrary heatmap, the downstream gains could come from a generic spatial reweighting rather than from semantic grounding.
  3. [Table 4, Sec. 4.4] The ablation evidence does not support the claim that the mask and text semantics drive the improvements. Removing the MGCA module (row a) improves VIF by 0.006 and Q_AB/F by 0.002 relative to the full model, and all ablation differences are within a few hundredths on most metrics. No error bars or multiple-run statistics are reported. The paper should add ablations with a random mask, a constant mask, and a mask-free variant that still performs spatial attention, so that the specific contribution of semantic grounding can be disentangled from the presence of any spatial weighting.
  4. [Sec. 4.3, Table 2] The downstream detection claims are stronger than the table supports. The text says TeSG is superior under mAP@0.75 and mAP@0.5:0.95, but in Table 2 the CDDFuse row reports 0.881 on the last column while the Ours row reports 0.879, and the mAP lead over CDDFuse is only 0.003. The column alignment also appears inconsistent with the header, which lists six metric columns while each row contains five numeric entries. Please clarify the exact metrics, correct the table, and temper the claims accordingly.
minor comments (6)
  1. [Fig. 1 caption] The figure caption refers to "TeGF (Our Method)", which does not match the method name TeSG used elsewhere in the paper.
  2. [Sec. 3.3] The notation for foreground and background features is confusingly typeset: the text uses F^m_v and F^m_v for what should be F^m_v and F^{\bar m}_v, and similarly for the other terms. Please use distinct superscripts throughout.
  3. [Fig. 3] The figure caption mentions "Reference Text: T" and "Query Text: T_Q", but T_Q is never defined in the text; presumably it denotes the modified caption \tilde{T} without V*. Please define it.
  4. [References] References [1]-[4] appear to be placeholder entries with "FirstName LastName" and fixed page ranges; these must be replaced with the actual cited works.
  5. [Table 2] The table header lists six metric columns (mAP, Precision, Recall, @0.50, @0.75, @0.50:0.95) but each row only has five numerical values; please align the table body with the header.
  6. [Sec. 4.4] The ablation section reports numerical differences of about 0.001 in some metrics without any measure of variance; at least note whether the differences are stable across training runs or report standard deviations.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: TeSG's textual and mask guidance is computed from the inputs and tested against independent fusion metrics and downstream detectors, so the central claim is not forced by construction.

full rationale

The derivation chain is self-contained rather than circular. The mask semantics M are generated from the input visible and infrared images in Eqs. (2)-(3) using a frozen BLIP caption and a frozen DDIM noise-difference, and the text semantics Ft are text embeddings of the same caption; these quantities are inputs to the MGCA and TDAF modules, while the fused image is trained with standard intensity, gradient, SSIM, and color losses. Nothing in the central claim is defined in terms of the fused output or the downstream metrics: the mask is not fitted to the YOLOv8 or DeeplabV3+ evaluators, and the reported gains are measured on external fusion metrics (EN, SD, SCD, VIF, Q_AB/F) and on independently trained downstream detectors and segmenters. The SIG pipeline is underspecified (keyword V* selection, binarization threshold, and infrared conditioning of the RGB diffusion model are not stated, and no mask-localization validation is provided), and the reference list contains placeholder entries that are not load-bearing, but these are reproducibility and quality gaps rather than circular steps. No equation reduces to its own output, no fitted parameter is renamed as a prediction, and no load-bearing self-citation chain is present.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the reliability of BLIP captions, the localization property of the DDIM noise difference, the compatibility of the diffusion model with infrared input, and the alignment of text embeddings with visual features. The mask binarization threshold and the keyword selection rule are unstated free choices that can influence all downstream results.

free parameters (3)
  • Binarization threshold in Delta = unspecified
    Eq. (2) normalizes and binarizes the noise difference to produce M_vis; the threshold is not reported, and results may be sensitive to it.
  • Keyword V* selection rule = unspecified
    In Sec. 3.2, a keyword V* is removed from the BLIP caption to create bT; the paper does not state whether V* is chosen automatically or manually, which affects the mask.
  • DDIM sampling steps for mask generation = unspecified
    The noise-based mask in Eq. (2) requires running the DDIM sampler; number of steps and noise level are not given in the main text.
assumptions (4)
  • domain assumption BLIP captions accurately describe the scene content relevant for fusion
    Used in Sec. 3.2 to generate T; if captions miss thermal targets, masks and text semantics are incomplete.
  • ad hoc to paper The difference in DDIM noise predictions with and without keyword V* localizes the keyword's region in the image
    Eq. (2) defines M_vis from this difference; no validation is provided for this localization property.
  • domain assumption The infrared image can be processed by the same diffusion model as the visible image
    Eq. (3) applies the same procedure to I_ir, which is single-channel; the paper does not specify how it is channel-expanded or whether the pretrained model accepts it.
  • domain assumption Text embeddings from the diffusion model's text encoder align with visual features for gating
    Assumed in Sec. 3.4 Eq. (7); no quantitative alignment check.

how reviews work

0 comments
Cite this review

Pith. "Pith review of TeSG: Textual Semantic Guidance for Infrared and Visible Image Fusion." pith.science (2026). https://pith.science/paper/F6OHY62F

@misc{pith2026250616730,
  author       = {Pith},
  title        = {Pith review of: TeSG: Textual Semantic Guidance for Infrared and Visible Image Fusion},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/F6OHY62F}},
  note         = {Machine review of arXiv:2506.16730}
}
read the original abstract

Infrared and visible image fusion (IVF) aims to combine complementary information from both image modalities, producing more informative and comprehensive outputs. Recently, text-guided IVF has shown great potential due to its flexibility and versatility. However, the effective integration and utilization of textual semantic information remains insufficiently studied. To tackle these challenges, we introduce textual semantics at two levels: the mask semantic level and the text semantic level, both derived from textual descriptions extracted by large Vision-Language Models (VLMs). Building on this, we propose Textual Semantic Guidance for infrared and visible image fusion, termed TeSG, which guides the image synthesis process in a way that is optimized for downstream tasks such as detection and segmentation. Specifically, TeSG consists of three core components: a Semantic Information Generator (SIG), a Mask-Guided Cross-Attention (MGCA) module, and a Text-Driven Attentional Fusion (TDAF) module. The SIG generates mask and text semantics based on textual descriptions. The MGCA module performs initial attention-based fusion of visual features from both infrared and visible images, guided by mask semantics. Finally, the TDAF module refines the fusion process with gated attention driven by text semantics. Extensive experiments demonstrate the competitiveness of our approach, particularly in terms of performance on downstream tasks, compared to existing state-of-the-art methods.

Figures

Figures reproduced from arXiv: 2506.16730 by the authors.

Figure 1
Figure 1. Comparison of existing IVF methods with our TeSG. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The overall framework of TeSG consists of three primary modules: (a) the Semantic Information Generator (SIG), responsible [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. The process of the Semantic Information Generator. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Qualitative comparison of our method against nine state-of-the-art (SOTA) methods. The images, arranged from top to bottom, [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Qualitative comparison of object detection performance on the MSRS dataset. [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Qualitative comparison of semantic segmentation performance on the MSRS dataset. [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]

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. Right Answer, Wrong Heat: Explanation-Aware Evaluation and Thermal-Grounded Feedback for MLLMs on Infrared Images

    cs.CV 2026-08 conditional novelty 6.0 of 10

    Many correct answers from vision-language models on infrared images are not grounded in thermal evidence, and a training-free feedback loop can repair the explanations.

Reference graph

Works this paper leans on

59 extracted references · 51 canonical work pages · cited by 1 Pith paper

  1. [1]

    Frobnication.IEEE TPAMI, 12(1):234– 778, 2002

    FirstName Alpher. Frobnication.IEEE TPAMI, 12(1):234– 778, 2002

    work page 2002

  2. [2]

    Frobnication revisited.Journal of Foo, 13(1):234–778, 2003

    FirstName Alpher and FirstName Fotheringham-Smythe. Frobnication revisited.Journal of Foo, 13(1):234–778, 2003

    work page 2003

  3. [3]

    Can a computer frobnicate? InCVPR, pages 234–778, 2005

    FirstName Alpher and FirstName Gamow. Can a computer frobnicate? InCVPR, pages 234–778, 2005

    work page 2005

  4. [4]

    Can a machine frobnicate?Journal of Foo, 14(1):234–778, 2004

    FirstName Alpher, FirstName Fotheringham-Smythe, and FirstName Gamow. Can a machine frobnicate?Journal of Foo, 14(1):234–778, 2004

    work page 2004

  5. [5]

    A new image quality metric for image fusion: The sum of the correlations of differences.Aeu-International Journal of Electronics and Communications, 69(12):1890– 1896, 2015

    V Aslantas. A new image quality metric for image fusion: The sum of the correlations of differences.Aeu-International Journal of Electronics and Communications, 69(12):1890– 1896, 2015. 5

    work page 2015

  6. [6]

    Two-scale im- age fusion of visible and infrared images using saliency de- tection.Infrared Physics & Technology, 76:52–64, 2016

    Durga Prasad Bavirisetti and Ravindra Dhuli. Two-scale im- age fusion of visible and infrared images using saliency de- tection.Infrared Physics & Technology, 76:52–64, 2016. 2

    work page 2016

  7. [7]

    Encoder-decoder with atrous separable convolution for semantic image segmentation

    Liang-Chieh Chen, Yukun Zhu, George Papandreou, Florian Schroff, and Hartwig Adam. Encoder-decoder with atrous separable convolution for semantic image segmentation. In Proceedings of the European Conference on Computer Vi- sion, pages 801–818, 2018. 7

    work page 2018

  8. [8]

    Textfusion: Unveiling the power of textual semantics for controllable image fusion

    Chunyang Cheng, Tianyang Xu, Xiao-Jun Wu, Hui Li, Xi Li, Zhangyong Tang, and Josef Kittler. Textfusion: Unveiling the power of textual semantics for controllable image fusion. Information Fusion, 117:102790, 2025. 2, 3, 5

    work page 2025

Show all 59 references
  1. [9]

    Region-based multimodal image fusion using ica bases

    Nedeljko Cvejic, David Bull, and Nishan Canagarajah. Region-based multimodal image fusion using ica bases. IEEE Sensors Journal, 7(5):743–751, 2007. 2

    2007

  2. [10]

    Background-subtraction using contour-based fusion of thermal and visible imagery

    James W Davis and Vinay Sharma. Background-subtraction using contour-based fusion of thermal and visible imagery. Computer Vision and Image Understanding, 106(2-3):162– 182, 2007. 1

    2007

  3. [11]

    Mmdrfuse: Distilled mini-model with dynamic refresh for multi-modality image fusion

    Yanglin Deng, Tianyang Xu, Chunyang Cheng, Xiao-Jun Wu, and Josef Kittler. Mmdrfuse: Distilled mini-model with dynamic refresh for multi-modality image fusion. InPro- ceedings of the 32nd ACM International Conference on Mul- timedia, pages 7326–7335, 2024. 5

    2024

  4. [12]

    Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio

    Ian J. Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial nets. InProceedings of the 28th International Conference on Neural Information Processing Systems, page 2672–2680, 2014. 2

    2014

  5. [13]

    Fast saliency-aware multi-modality image fusion.Neurocomput- ing, 111:70–80, 2013

    Jungong Han, Eric J Pauwels, and Paul De Zeeuw. Fast saliency-aware multi-modality image fusion.Neurocomput- ing, 111:70–80, 2013. 1

    2013

  6. [14]

    A new im- age fusion performance metric based on visual information fidelity.Information fusion, 14(2):127–135, 2013

    Yu Han, Yunze Cai, Yin Cao, and Xiaoming Xu. A new im- age fusion performance metric based on visual information fidelity.Information fusion, 14(2):127–135, 2013. 5

    2013

  7. [15]

    Reconet: Recurrent correction network for fast and efficient multi-modality image fusion

    Zhanbo Huang, Jinyuan Liu, Xin Fan, Risheng Liu, Wei Zhong, and Zhongxuan Luo. Reconet: Recurrent correction network for fast and efficient multi-modality image fusion. InProceedings of the European Conference on Computer Vi- sion, pages 539–555, 2022. 5

    2022

  8. [16]

    Llvip: A visible-infrared paired dataset for low-light vision

    Xinyu Jia, Chuang Zhu, Minzhen Li, Wenqi Tang, and Wenli Zhou. Llvip: A visible-infrared paired dataset for low-light vision. InProceedings of the IEEE/CVF International Con- ference on Computer Vision, pages 3496–3504, 2021. 5, 6, 8

    2021

  9. [17]

    Weiwei Kong, Yang Lei, and Huaixun Zhao. Adaptive fusion method of visible light and infrared images based on non- subsampled shearlet transform and fast non-negative matrix factorization.Infrared Physics & Technology, 67:161–172, 2014

    2014

  10. [18]

    The frobnicatable foo filter, 2014

    FirstName LastName. The frobnicatable foo filter, 2014. Face and Gesture submission ID 324. Supplied as supple- mental materialfg324.pdf

    2014

  11. [19]

    Frobnication tutorial, 2014

    FirstName LastName. Frobnication tutorial, 2014. Supplied as supplemental materialtr.pdf

    2014

  12. [20]

    An infrared and visible image fusion method based on multi-scale transfor- mation and norm optimization.Information Fusion, 71:109– 129, 2021

    Guofa Li, Yongjie Lin, and Xingda Qu. An infrared and visible image fusion method based on multi-scale transfor- mation and norm optimization.Information Fusion, 71:109– 129, 2021. 2

    2021

  13. [21]

    Densefuse: A fusion approach to infrared and visible images.IEEE Transactions on Image Processing, 28(5):2614–2623, 2018

    Hui Li and Xiao-Jun Wu. Densefuse: A fusion approach to infrared and visible images.IEEE Transactions on Image Processing, 28(5):2614–2623, 2018. 2

    2018

  14. [22]

    An improved fusion algorithm for infrared and visible images based on multi-scale transform.Infrared Physics & Technology, 74: 28–37, 2016

    He Li, Lei Liu, Wei Huang, and Chao Yue. An improved fusion algorithm for infrared and visible images based on multi-scale transform.Infrared Physics & Technology, 74: 28–37, 2016. 2

    2016

  15. [23]

    Image registration and fusion of visible and infrared inte- grated camera for medium-altitude unmanned aerial vehicle remote sensing.Remote Sensing, 9(5):441, 2017

    Hongguang Li, Wenrui Ding, Xianbin Cao, and Chunlei Liu. Image registration and fusion of visible and infrared inte- grated camera for medium-altitude unmanned aerial vehicle remote sensing.Remote Sensing, 9(5):441, 2017. 1

    2017

  16. [24]

    Rfn-nest: An end-to- end residual fusion network for infrared and visible images

    Hui Li, Xiao-Jun Wu, and Josef Kittler. Rfn-nest: An end-to- end residual fusion network for infrared and visible images. Information Fusion, 73:72–86, 2021. 2

    2021

  17. [25]

    Hui Li, Tianyang Xu, Xiao-Jun Wu, Jiwen Lu, and Josef Kit- tler. Lrrnet: A novel representation learning guided fusion network for infrared and visible images.IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(9):11040– 11052, 2023. 5

    2023

  18. [26]

    Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation

    Junnan Li, Dongxu Li, Caiming Xiong, and Steven Hoi. Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation. InProceed- ings of the 39th International Conference on Machine Learn- ing, pages 12888–12900. PMLR, 2022. 2, 3, 4

    2022

  19. [27]

    Image fusion with guided filtering.IEEE Transactions on Image Process- ing, 22(7):2864–2875, 2013

    Shutao Li, Xudong Kang, and Jianwen Hu. Image fusion with guided filtering.IEEE Transactions on Image Process- ing, 22(7):2864–2875, 2013. 2

    2013

  20. [28]

    Infrared and visible image fusion method based on saliency detection in sparse domain

    CH Liu, Yue Qi, and WR Ding. Infrared and visible image fusion method based on saliency detection in sparse domain. Infrared Physics & Technology, 83:94–102, 2017. 2

    2017

  21. [29]

    Learning a deep multi-scale feature ensemble and an edge-attention guidance for image fusion.IEEE Transac- tions on Circuits and Systems for Video Technology, 32(1): 105–119, 2021

    Jinyuan Liu, Xin Fan, Ji Jiang, Risheng Liu, and Zhongx- uan Luo. Learning a deep multi-scale feature ensemble and an edge-attention guidance for image fusion.IEEE Transac- tions on Circuits and Systems for Video Technology, 32(1): 105–119, 2021. 2

    2021

  22. [30]

    Target-aware dual adversarial learning and a multi-scenario multi-modality benchmark to fuse infrared and visible for object detection

    Jinyuan Liu, Xin Fan, Zhanbo Huang, Guanyao Wu, Risheng Liu, Wei Zhong, and Zhongxuan Luo. Target-aware dual adversarial learning and a multi-scenario multi-modality benchmark to fuse infrared and visible for object detection. InProceedings of the IEEE/CVF Conference on Comput...

    2022

  23. [31]

    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. InProceedings of the IEEE/CVF International Conference on Computer Visi...

    2023

  24. [32]

    Infrared and visible image fusion with convolutional neural networks.International Journal of Wavelets, Multiresolution and Information Processing, 16(03):1850018, 2018

    Yu Liu, Xun Chen, Juan Cheng, Hu Peng, and Zengfu Wang. Infrared and visible image fusion with convolutional neural networks.International Journal of Wavelets, Multiresolution and Information Processing, 16(03):1850018, 2018. 2

    2018

  25. [33]

    In- frared and visible image fusion based on visual saliency map and weighted least square optimization.Infrared Physics & Technology, 82:8–17, 2017

    Jinlei Ma, Zhiqiang Zhou, Bo Wang, and Hua Zong. In- frared and visible image fusion based on visual saliency map and weighted least square optimization.Infrared Physics & Technology, 82:8–17, 2017. 2

    2017

  26. [34]

    Infrared and visible im- age fusion methods and applications: A survey.Information fusion, 45:153–178, 2019

    Jiayi Ma, Yong Ma, and Chang Li. Infrared and visible im- age fusion methods and applications: A survey.Information fusion, 45:153–178, 2019. 1, 5, 6

    2019

  27. [35]

    Fusiongan: A generative adversarial network for in- frared and visible image fusion.Information fusion, 48:11– 26, 2019

    Jiayi Ma, Wei Yu, Pengwei Liang, Chang Li, and Junjun Jiang. Fusiongan: A generative adversarial network for in- frared and visible image fusion.Information fusion, 48:11– 26, 2019. 2

    2019

  28. [36]

    Infrared and visible image fusion via detail preserving adversarial learning.Information Fu- sion, 54:85–98, 2020

    Jiayi Ma, Pengwei Liang, Wei Yu, Chen Chen, Xiaojie Guo, Jia Wu, and Junjun Jiang. Infrared and visible image fusion via detail preserving adversarial learning.Information Fu- sion, 54:85–98, 2020

    2020

  29. [37]

    Ddcgan: A dual-discriminator conditional gen- erative adversarial network for multi-resolution image fu- sion.IEEE Transactions on Image Processing, 29:4980– 4995, 2020

    Jiayi Ma, Han Xu, Junjun Jiang, Xiaoguang Mei, and Xiao- Ping Zhang. Ddcgan: A dual-discriminator conditional gen- erative adversarial network for multi-resolution image fu- sion.IEEE Transactions on Image Processing, 29:4980– 4995, 2020. 2

    2020

  30. [38]

    Jiayi Ma, Hao Zhang, Zhenfeng Shao, Pengwei Liang, and Han Xu. Ganmcc: A generative adversarial network with multiclassification constraints for infrared and visible image fusion.IEEE Transactions on Instrumentation and Measure- ment, 70:1–14, 2020. 2

    2020

  31. [39]

    Swinfusion: Cross-domain long-range learning for general image fusion via swin transformer

    Jiayi Ma, Linfeng Tang, Fan Fan, Jun Huang, Xiaoguang Mei, and Yong Ma. Swinfusion: Cross-domain long-range learning for general image fusion via swin transformer. IEEE/CAA Journal of Automatica Sinica, 9(7):1200–1217,

  32. [40]

    Conditional generative adversarial nets.arXiv preprint arXiv:1411.1784, 2014

    Mehdi Mirza and Simon Osindero. Conditional generative adversarial nets.arXiv preprint arXiv:1411.1784, 2014. 2

    2014 arXiv

  33. [41]

    Denois- ing diffusion implicit models

    Jiaming Song, Chenlin Meng, and Stefano Ermon. Denois- ing diffusion implicit models. InInternational Conference on Learning Representations, 2021. 4

    2021

  34. [42]

    Image fusion in the loop of high-level vision tasks: A semantic-aware real- time infrared and visible image fusion network.Information Fusion, 82:28–42, 2022

    Linfeng Tang, Jiteng Yuan, and Jiayi Ma. Image fusion in the loop of high-level vision tasks: A semantic-aware real- time infrared and visible image fusion network.Information Fusion, 82:28–42, 2022. 3

    2022

  35. [43]

    Piafusion: A progressive infrared and visible im- age fusion network based on illumination aware.Information Fusion, 83:79–92, 2022

    Linfeng Tang, Jiteng Yuan, Hao Zhang, Xingyu Jiang, and Jiayi Ma. Piafusion: A progressive infrared and visible im- age fusion network based on illumination aware.Information Fusion, 83:79–92, 2022. 5, 6, 7

    2022

  36. [44]

    Terf: Text-driven and region- aware flexible visible and infrared image fusion

    Hebaixu Wang, Hao Zhang, Xunpeng Yi, Xinyu Xiang, Leyuan Fang, and Jiayi Ma. Terf: Text-driven and region- aware flexible visible and infrared image fusion. InProceed- ings of the 32nd ACM International Conference on Multime- dia, pages 935–944, 2024. 2, 3

    2024

  37. [45]

    Fusion method for infrared and visible images by using non-negative sparse representation.Infrared Physics & Technology, 67:477–489, 2014

    Jun Wang, Jinye Peng, Xiaoyi Feng, Guiqing He, and Jian- ping Fan. Fusion method for infrared and visible images by using non-negative sparse representation.Infrared Physics & Technology, 67:477–489, 2014. 2

    2014

  38. [46]

    Image quality assessment: from error visibility to structural similarity.IEEE Transactions on Image Process- ing, 13(4):600–612, 2004

    Zhou Wang, Alan C Bovik, Hamid R Sheikh, and Eero P Si- moncelli. Image quality assessment: from error visibility to structural similarity.IEEE Transactions on Image Process- ing, 13(4):600–612, 2004

    2004

  39. [47]

    Swinfuse: A residual swin transformer fusion net- work for infrared and visible images.IEEE Transactions on Instrumentation and Measurement, 71:1–12, 2022

    Zhishe Wang, Yanlin Chen, Wenyu Shao, Hui Li, and Lei Zhang. Swinfuse: A residual swin transformer fusion net- work for infrared and visible images.IEEE Transactions on Instrumentation and Measurement, 71:1–12, 2022. 2

    2022

  40. [48]

    U2fusion: A unified unsupervised image fusion net- work.IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(1):502–518, 2020

    Han Xu, Jiayi Ma, Junjun Jiang, Xiaojie Guo, and Haibin Ling. U2fusion: A unified unsupervised image fusion net- work.IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(1):502–518, 2020. 2, 5, 6

    2020

  41. [49]

    Rfnet: Unsupervised network for mutually reinforcing multi- modal image registration and fusion

    Han Xu, Jiayi Ma, Jiteng Yuan, Zhuliang Le, and Wei Liu. Rfnet: Unsupervised network for mutually reinforcing multi- modal image registration and fusion. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 19679–19688, 2022. 2

    2022

  42. [50]

    Multimodal information fusion for urban scene understanding.Machine Vision and Applica- tions, 27:331–349, 2016

    Philippe Xu, Franck Davoine, Jean-Baptiste Bordes, Huijing Zhao, and Thierry Denœux. Multimodal information fusion for urban scene understanding.Machine Vision and Applica- tions, 27:331–349, 2016. 1

    2016

  43. [51]

    Text-if: Leveraging semantic text guidance for degradation-aware and interactive image fusion

    Xunpeng Yi, Han Xu, Hao Zhang, Linfeng Tang, and Ji- ayi Ma. Text-if: Leveraging semantic text guidance for degradation-aware and interactive image fusion. InProceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 27026–27035, 2024. 2, 3, 5

    2024

  44. [52]

    Dif-fusion: Toward high color fidelity in infrared and visible image fusion with diffusion models.IEEE Transactions on Image Processing, 32:5705–5720, 2023

    Jun Yue, Leyuan Fang, Shaobo Xia, Yue Deng, and Jiayi Ma. Dif-fusion: Toward high color fidelity in infrared and visible image fusion with diffusion models.IEEE Transactions on Image Processing, 32:5705–5720, 2023. 2

    2023

  45. [53]

    Sdnet: A versatile squeeze-and- decomposition network for real-time image fusion.Inter- national Journal of Computer Vision, 129(10):2761–2785,

    Hao Zhang and Jiayi Ma. Sdnet: A versatile squeeze-and- decomposition network for real-time image fusion.Inter- national Journal of Computer Vision, 129(10):2761–2785,

  46. [54]

    Image fusion meets deep learning: A survey and perspective

    Hao Zhang, Han Xu, Xin Tian, Junjun Jiang, and Jiayi Ma. Image fusion meets deep learning: A survey and perspective. Information Fusion, 76:323–336, 2021. 1

    2021

  47. [55]

    Metafusion: Infrared and visible image fusion via meta- feature embedding from object detection

    Wenda Zhao, Shigeng Xie, Fan Zhao, You He, and Huchuan Lu. Metafusion: Infrared and visible image fusion via meta- feature embedding from object detection. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13955–13965, 2023. 5

    2023

  48. [56]

    Didfuse: deep image de- composition for infrared and visible image fusion

    Zixiang Zhao, Shuang Xu, Chunxia Zhang, Junmin Liu, Jiangshe Zhang, and Pengfei Li. Didfuse: deep image de- composition for infrared and visible image fusion. InPro- ceedings of the 29th International Conference on Interna- tional Joint Conferences on Artificial Intelligence, ...

    2021

  49. [57]

    Cddfuse: Correlation-driven dual-branch feature decompo- sition for multi-modality image fusion

    Zixiang Zhao, Haowen Bai, Jiangshe Zhang, Yulun Zhang, Shuang Xu, Zudi Lin, Radu Timofte, and Luc Van Gool. Cddfuse: Correlation-driven dual-branch feature decompo- sition for multi-modality image fusion. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern ...

    2023

  50. [58]

    Ddfm: denoising diffusion model for multi-modality image fusion

    Zixiang Zhao, Haowen Bai, Yuanzhi Zhu, Jiangshe Zhang, Shuang Xu, Yulun Zhang, Kai Zhang, Deyu Meng, Radu Timofte, and Luc Van Gool. Ddfm: denoising diffusion model for multi-modality image fusion. InProceedings of the IEEE/CVF International Conference on Computer Vi- sion, pa...

    2023

  51. [59]

    Image fusion via vision-language model

    Zixiang Zhao, Lilun Deng, Haowen Bai, Yukun Cui, Zhipeng Zhang, Yulun Zhang, Haotong Qin, Dongdong Chen, Jiangshe Zhang, Peng Wang, et al. Image fusion via vision-language model. InProceedings of the 41st In- ternational Conference on Machine Learning, pages 60749– 60765, 2024. 2, 3

    2024

Pith tools

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