Pith. sign in

REVIEW 3 major objections 4 minor 145 references

CFGPNet: Cross-Attention-Based Fused Gradient Programmed Network Framework for Multispectral Object Detection

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 ↗

Pith's one-line read CFGPNet claims that visible–thermal object detection reaches leading results on five benchmarks by having each modality re-weight the other through exchanged spatial attention maps, with a gradient-programmed auxiliary branch carrying the…

desk verdict A detailed, transparent architecture paper whose headline numbers are undermined by an internal contradiction between the specified final loss and the ablation table; the reported SOTA is not reproducible until that is resolved. read the letter →

arxiv 2608.06205 v1 pith:BD36R2GG submitted 2026-08-06 cs.CV

Nima Hatami , Karim Faez , Saeed Sharifian , Hamidreza Amindavar This is my paper
classification cs.CV
keywords multispectralobjectdetectioncross-modalfeaturefusioninteractionprogrammablegradientinformationcross-attentionRGB-Tvisible-infraredre-parameterizedbackbone
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 visible-plus-thermal object detection works best when the two modalities cooperate through exchanged spatial attention maps rather than raw features: each stream computes a compact reliability map and the other stream is re-weighted by it, preserving texture and thermal evidence while suppressing single-modality bias. Around that idea the paper assembles a full detector, CFGPNet, using a re-parameterized YOLOv9-style backbone, a selective aggregation fusion module, and a gradient-programming auxiliary branch that keeps a second detection path alive during training and inference. CFGPNet reports state-of-the-art or leading results on all five public RGB–T benchmarks it evaluates, including 80.7% mAP50 / 45.0% mAP50:95 on FLIR and 83.4% / 61.8% on MFAD, across three model scales from 21.0M to 180.9M parameters. A sympathetic reader would care because low-light and adverse-weather perception for driving, surveillance, and robotics depends on exactly this kind of fusion, and the paper argues it can be done without the heavy cost of dense Transformer-style cross-attention.

What carries the argument

The load-bearing mechanism is CrossCEA, a cross-attention module that swaps compact spatial reliability maps between the visible and thermal branches. Inside each CEA unit, the input feature map is split into channel groups; axis-pooled descriptors pass through depthwise 1D convolutions, group normalization, and sigmoid to form height- and width-wise gates, while a parallel global branch builds channel queries; a temperature-scaled interaction between the local and global paths produces a single 2D attention map $\Omega$ per modality. CrossCEA then gates each stream with the other stream's map, $\tilde{F}^I = F^I \odot \Omega^V$ and $\tilde{F}^V = F^V \odot \Omega^I$, so that only bounded attention weights, not raw feature values, cross between modalities. The second mechanism is ASAF, which concatenates the enhanced streams and runs two parallel paths — a DenseNet-style dense aggregation refined by CBAM, and an ELA-guided multi-branch selector (MBatt) whose element-wise max keeps the strongest attention candidate at each location — before a 1×1 projection yields the compact fused tensor. The third mechanism is optimization-level: the PGI auxiliary branch instantiates a second detection path whose loss (weighted 0.25) sends programmed gradients back through tapped intermediate features of the main branch, and the branch stays active at inference.

What would settle it

Retrain the strongest published RGB–T baselines under CFGPNet's exact protocol — 600 epochs from scratch at 640×640, the same batch-size-adjusted schedules, the same dataset-specific NMS IoU thresholds chosen on 10% held-out validation subsets, and best-checkpoint selection — and compare on the five benchmarks. If a baseline matches or exceeds CFGPNet under that protocol, the architecture-attributed gains are refuted; a complementary check replaces CrossCEA's swapped attention maps with concatenation or early feature mixing at matched computation and asks whether the mAP50:95 margin survives.

Watch

Extended reading notes

Core claim

The paper's central claim is that a multispectral detector built from three interacting mechanisms — a RepViT-block GELAN backbone with separate visible and infrared weights, a Cross Computation Efficient Attention (CrossCEA) stage that exchanges compact spatial attention maps between the two streams, and an Attention Selection and Aggregation Fusion (ASAF) network that condenses the enhanced streams into one tensor — outperforms recent RGB–T detectors on every benchmark it evaluates. The best reported configurations reach 80.7% mAP50 / 45.0% mAP50:95 on FLIR, 89.9% / 63.4% on M3FD, 97.8% / 68.9% on LLVIP, 83.3% / 56.9% on VEDAI, and 83.4% / 61.8% on MFAD. The author would add that the framework is trained from scratch in three scales (21.0M, 71.7M, and 180.9M parameters), that the programmable-gradient auxiliary branch remains active at inference and contributes candidate detections merged by non-maximum suppression, and that the MFAD ablations attribute the gains to the combination of the RepViT backbone, CrossCEA, ASAF, and the PGI branch rather than to any single module.

Load-bearing premise

The reported superiority over prior detectors assumes that CFGPNet's longer training schedule (600 epochs from scratch), per-dataset NMS IoU threshold tuning, and best-checkpoint selection are not what produce the gains, since no published baseline is retrained under the same protocol.

Editorial extensions

If this is right

  • If the reported results hold, the 'exchange attention maps, not features' principle offers a lightweight alternative to dense Transformer cross-attention for RGB–T fusion, since CrossCEA adds no token-wise attention and the three scales report 18.6–52.2 FPS at 640×640 input.
  • The compact 21.0M-parameter variant claims to match or beat much larger published fusion detectors on FLIR, M3FD, and LLVIP, suggesting strong parameter-efficient RGB–T detection is possible without ImageNet-pretrained initialization.
  • The MFAD ablation quantifies the optimization contribution: removing the PGI branch drops mAP50 from 79.6 to 74.1 and mAP50:95 from 56.7 to 53.3, so the gradient-programming path is a substantial part of the reported gain.
  • The paper's VEDAI evaluation uses Split 1 for all variants and includes only single-split reports from prior work, so its VEDAI comparisons are conditional on that split being representative of the benchmark.

Reading between the lines

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

  • The paper never runs the clean attribution experiment: retraining strong published baselines under its exact 600-epoch, from-scratch protocol with identical NMS tuning and checkpoint selection; without that control, the size of the architecture-level margin over prior work is an open question.
  • The attention-map-swap principle is a transferable hypothesis for any aligned sensor pair with location-dependent reliability — RGB-depth, RGB-event, or multi-spectral satellite imagery — and could be tested at matched compute against concatenation, learned-gate, or Transformer fusion.
  • The ablations show that pointwise and structural alignment losses (MSE, SSIM) actively degrade accuracy, which suggests forcing modalities toward one another is counterproductive; a useful follow-up is to test whether any cross-modal alignment term is needed once the backbone and fusion are strong.
  • The abstract's LLVIP headline pair mixes scales: 97.8% mAP50 comes from CFGPNet-m and 68.9% mAP50:95 from CFGPNet-e, and no single variant achieves both; with accuracy reported as single-run point estimates, small margins on saturated benchmarks may fall within run-to-run variance, which the conclusion itself flags as future work.
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

3 major / 4 minor

Summary. The paper proposes CFGPNet, a multispectral (RGB-T) object detection framework combining a RepViT-based GELAN backbone, a CrossCEA cross-modal attention module, an ASAF fusion module, and a programmable-gradient auxiliary branch (PGI). Three model scales (m/c/e) are designed and evaluated on FLIR, M3FD, LLVIP, VEDAI, and MFAD, where the paper reports state-of-the-art or near-state-of-the-art results. The authors also provide detailed layer-wise configuration tables, release code/data/models, and conduct ablation studies on MFAD to support the contribution of individual components.

Significance. If the reported results are reproducible, CFGPNet would represent a practical advance in multispectral detection, combining competitive accuracy with a favorable efficiency trade-off. The paper is commendable for its detailed architectural descriptions, the release of code and fine-tuned models, and the broad evaluation across five benchmarks with per-class breakdowns. The ablations are extensive and mostly controlled, and the disclosure of dataset-specific NMS thresholds and the use of a single VEDAI split is transparent. However, the current manuscript contains a load-bearing internal contradiction in the training-loss specification that prevents verification of which configuration produced the headline results; this issue, together with a lack of same-protocol baselines, makes the central claim not yet defensible as written.

major comments (3)
  1. [Section 3.6 vs. Table 12] The training objective is specified inconsistently. In Section 3.6, the total loss is defined as the YOLOv9 detection losses plus all three proposed alignment terms, with λ_CE−KL=1.0, λ_SSIM=3.0, and λ_MSE=300.0, and the text states that these coefficients are kept fixed across datasets. However, Table 12 of the ablation study reports that this exact combination (MSE+SSIM+CE-KL) yields only 74.3 mAP50 / 51.5 mAP50:95 on MFAD, while CFGPNet-m is reported at 79.6 / 56.7. The same table identifies CE-KL alone as the best loss, and the text in Section 4.5 explicitly states that "CE-KL is selected for the final model." These two statements are mutually contradictory. If the final models were trained with the Section 3.6 loss, the headline results are contradicted by the paper's own ablation; if the final models used CE-KL only, then Section 3.6 and its hyperparameters do not describe the actual training objective. Since the loss function is central to the optimization, the reported state-of-the-art numbers cannot be attributed to the proposed architecture, and independent reproduction is blocked. The authors must correct this inconsistency and specify exactly which loss was used for each reported result, ideally by providing training logs or rerunning the affected experiments.
  2. [Sections 4.3 and 4.4] The comparison with prior methods is not conducted under equal conditions. All CFGPNet variants are trained from scratch for 600 epochs, use dataset-specific NMS IoU thresholds selected on a 10% validation subset (Table 2), and are evaluated with the best checkpoint over the training schedule (as stated for VEDAI). In contrast, baseline results are copied from their original publications, which may use different training budgets, data splits, input resolutions, and post-processing settings. Consequently, the reported gains cannot be isolated to the architectural contributions (CrossCEA, ASAF, PGI). To support the claim of architecture-level superiority, the authors should provide at least one strong baseline (e.g., a YOLOv9-based dual-stream detector, or the best performing published method) trained under the same 600-epoch schedule, same data augmentation, same NMS calibration, and same checkpoint-selection protocol. Without such a control, the state-of-the-art claim is not convincingly established.
  3. [Table 12 and Section 3.6] Even if the loss-contradiction is resolved, the ablation provides only weak support for the proposed feature-alignment loss. In Table 12, the "None" configuration achieves 79.7 mAP50 and 56.6 mAP50:95, while CE-KL alone achieves 79.6 and 56.7, i.e., a 0.1-point improvement on mAP50:95 and a 0.1-point drop on mAP50. The text acknowledges this but nevertheless selects CE-KL. Given that one of the paper's stated contributions is the auxiliary feature-alignment supervision of CrossCEA, the negligible and sometimes negative effect of the alignment loss should be discussed more candidly. The authors should either provide additional ablations (e.g., different λ values, other datasets) showing a consistent benefit, or moderate the claim that this loss is a key enabler of the reported performance.
minor comments (4)
  1. [Table 14 / Section 4.5] The MBatt branch ablation reports results for 2 to 6 branches, but the text refers to the case of a single branch ("When only one branch is used..."); either add the N_b=1 row to the table or adjust the discussion to avoid referencing a configuration that is not tabulated.
  2. [Tables 3 vs 10–16] The FPS values differ by an order of magnitude between the main comparison (e.g., CFGPNet-m: 52.2 FPS) and the ablation tables (e.g., 1463.4 FPS for the same model). Although the batched timing protocol is explained, the presentation is confusing. Please clearly label ablation FPS as "relative batched timing" in every table caption and in the text wherever such values appear.
  3. [Section 4.4.4 (VEDAI)] The sentence "each variant is evaluated using its best-performing checkpoint over the same 600-epoch training schedule" should specify whether the best checkpoint is chosen using the validation subset or the test split. If the test set was used for checkpoint selection, the reported numbers would be optimistically biased.
  4. [Abstract and Section 4.4] The claim of "strong and consistent performance" is weaker on FLIR, where the mAP50 gain over the best published method (ERFF, 80.6) is only 0.1 point for CFGPNet-e. Please present a more nuanced interpretation of the margin of improvement on each dataset.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the reported gains are empirical benchmark measurements, and the loss-function inconsistency is a reproducibility flaw rather than a derivation that reduces to its inputs.

full rationale

CFGPNet's central claims are empirical detection scores on five public benchmarks produced by training the proposed architecture. There is no derivation chain in which a predicted quantity is constructed from the same quantity, and no fitted parameter is renamed as a prediction. The NMS IoU thresholds are tuned on a held-out 10% validation subset and then fixed before evaluation; this is standard post-processing calibration and does not make the reported mAP values equivalent to the tuning objective. The learnable aggregation factors are explicitly ablated and discarded, so they are not fitted inputs masquerading as predictions. The paper does not rely on load-bearing self-citations: references to YOLOv9, RepViT, EMA, ELA, DenseNet, and CBAM are external published methods, and no uniqueness theorem from the authors is invoked. The internal contradiction between §3.6, which specifies the final objective as YOLOv9 losses plus MSE, SSIM, and CE-KL with the stated coefficients, and Table 12, which shows that MSE+SSIM+CE-KL degrades CFGPNet-m to 74.3/51.5 and that CE-KL alone is selected, is a serious reproducibility and correctness flaw that prevents unambiguous attribution of the reported gains to the proposed modules. However, this is not circularity in the derivation sense: the reported numbers are still measurements on external benchmarks, and no equation reduces an output to an input by construction. Single-run point estimates and differences in training budgets between CFGPNet and published baselines are statistical fairness concerns, not circularity. Overall, the paper is self-contained as an empirical architecture study, so the circularity score is 0.

Assumptions & free parameters 9 free parameters · 4 assumptions · 0 invented entities

The central claims depend on validation-tuned hyperparameters (NMS thresholds, loss weights, branch count) and on the assumption that cited baseline numbers are fair and accurate. The architecture introduces no new physical entities; its new modules are computational components.

free parameters (9)
  • NMS IoU threshold (FLIR) = 0.6
    Selected on a held-out 10% validation subset per dataset; affects final detections after NMS.
  • NMS IoU threshold (M3FD) = 0.5
    Selected on validation subset, dataset-specific.
  • NMS IoU threshold (LLVIP) = 0.65
    Selected on validation subset.
  • NMS IoU threshold (VEDAI) = 0.45
    Selected on validation subset.
  • NMS IoU threshold (MFAD) = 0.55
    Selected on validation subset.
  • Loss weights lambda_CE-KL, lambda_SSIM, lambda_MSE = 1.0, 3.0, 300.0
    Declared in Sec. 3.6 as empirically selected scale-compensation coefficients; however, ablation Table 12 shows the combined terms degrade performance, so the actual values used in the final model are unclear.
  • rho_PGI (auxiliary branch weight) = 0.25
    Chosen to down-weight the auxiliary branch.
  • Detection loss gains (box, cls, DFL) = 7.5, 0.5, 1.5
    Follows YOLOv9-style loss balancing, not derived.
  • Number of MBatt branches Nb = 3
    Ablated as optimal; a hand-selected hyperparameter.
assumptions (4)
  • standard math Softmax and scaled dot-product attention operations are computed as defined (Eq. 14).
    Standard attention formulation used as background.
  • domain assumption The registered RGB-T pairs share the same field of view and are synchronized, so cross-modal attention map swapping is meaningful.
    Dataset descriptions in Sec. 4.1 state registered pairs; the method relies on this alignment.
  • domain assumption The reported results of compared methods were obtained under comparable protocols and are accurate as cited.
    The comparison tables adopt values from other publications without re-evaluation, assuming consistency.
  • ad hoc to paper The alignment losses (CE-KL, SSIM, MSE) improve fusion; however, the ablation shows these losses do not clearly help, making this an ad hoc assumption of the paper.
    Table 12 shows the full combination degrades performance; final-model configuration is ambiguous.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CFGPNet: Cross-Attention-Based Fused Gradient Programmed Network Framework for Multispectral Object Detection." pith.science (2026). https://pith.science/paper/BD36R2GG

@misc{pith2026260806205,
  author       = {Pith},
  title        = {Pith review of: CFGPNet: Cross-Attention-Based Fused Gradient Programmed Network Framework for Multispectral Object Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BD36R2GG}},
  note         = {Machine review of arXiv:2608.06205}
}
read the original abstract

RGB--T object detection exploits the complementary strengths of visible and infrared imagery, supporting robust perception in low-light, adverse-weather, and complex multi-scale environments. However, existing methods still suffer from insufficient cross-modal interaction, unstable fusion from modality distribution gaps, and the high computational cost of heavy attention-based architectures. To address these issues, CFGPNet is proposed, a Cross-Attention-Based Fused Gradient Programmed Network framework for multispectral object detection. CFGPNet uses an improved GELAN backbone with RepViT-style re-parameterized blocks to strengthen feature representation while preserving computational efficiency. A Cross Computation Efficient Attention (CrossCEA) module is introduced to enhance cross-modal feature interaction and reduce redundant information transfer between visible and thermal branches. To generate compact and discriminative fused representations, an Attention Selection and Aggregation Fusion (ASAF) network combines dense feature aggregation with selective attention-based emphasis. Moreover, a programmable-gradient auxiliary branch is integrated into each CFGPNet variant to improve gradient delivery and optimization quality. Experiments on five public multispectral benchmarks, FLIR, M3FD, LLVIP, VEDAI, and MFAD, demonstrate that CFGPNet achieves strong and consistent performance across diverse scenes, object scales, and modality balances. In particular, the framework attains 80.7% mAP50 / 45.0% mAP50:95 on FLIR, 89.9% / 63.4% on M3FD, and 97.8% / 68.9% on LLVIP. It also reaches 83.3% / 56.9% on VEDAI and 83.4% / 61.8% on MFAD. These results show that CFGPNet is an effective, practical solution offering useful accuracy--efficiency trade-offs across three model scales. The code, data, and fine-tuned models are available at https://github.com/NimaHatami99/CFGPNet.

Figures

Figures reproduced from arXiv: 2608.06205 by the authors.

Figure 1
Figure 1. Illustration of the compact model inside the CFGPNet framework, which contains the main and auxiliary branches in the backbone, the cross-attention [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Structure diagram of (a) RepViTCSPELAN4, (b) RepViTCSP, and (c) RepViTBottleneck (the two arguments [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Overall structure diagram of the MFE module. [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (14 more)
Figure 4
Figure 4. Figure 4: Illustration of structure diagrams (a) ELA, (b) EMA, and (c) CEA. The dimensions of the module outputs are shown in the figure. [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]
Figure 5
Figure 5. Figure 5: Overview of the ASAF module. The upper branch performs dense aggregation and CBAM-based refinement, whereas the lower MBatt branch generates [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: Illustration of the extended model inside the CFGPNet framework, with multi-level auxiliary information transferred from the auxiliary branch to separate [PITH_FULL_IMAGE:figures/full_fig_p016_6.png]
Figure 7
Figure 7. Figure 7: Comparison between CFGPNet and state-of-the-art models in terms [PITH_FULL_IMAGE:figures/full_fig_p020_7.png]
Figure 8
Figure 8. Figure 8: Comparison between CFGPNet and state-of-the-art models in terms [PITH_FULL_IMAGE:figures/full_fig_p021_8.png]
Figure 9
Figure 9. Figure 9: Comparison between CFGPNet and state-of-the-art models in terms of [PITH_FULL_IMAGE:figures/full_fig_p023_9.png]
Figure 10
Figure 10. Figure 10: Comparison between CFGPNet and state-of-the-art models in terms [PITH_FULL_IMAGE:figures/full_fig_p025_10.png]
Figure 11
Figure 11. Figure 11: Comparison between CFGPNet and state-of-the-art models in terms of [PITH_FULL_IMAGE:figures/full_fig_p026_11.png]
Figure 12
Figure 12. Figure 12: Illustration of precision–recall curves for CFGPNet-e on (a) FLIR, (b) M3FD, (c) LLVIP, (d) VEDAI, and (e) MFAD datasets. [PITH_FULL_IMAGE:figures/full_fig_p027_12.png]
Figure 13
Figure 13. Figure 13: Illustration of inference results for CFT, C [PITH_FULL_IMAGE:figures/full_fig_p031_13.png]
Figure 14
Figure 14. Figure 14: Illustration of inference results for CFT, DMFusion-YOLOv8, ICAFusion, EI [PITH_FULL_IMAGE:figures/full_fig_p033_14.png]
Figure 15
Figure 15. Figure 15: Illustration of inference results for CFT, MetaFusion, ICAFusion, COMO, and CFGPNet-e on the LLVIP dataset. The yellow arrow annotations show [PITH_FULL_IMAGE:figures/full_fig_p034_15.png]
Figure 16
Figure 16. Figure 16: Illustration of inference results for CFT, COMO, ICAFusion, [PITH_FULL_IMAGE:figures/full_fig_p035_16.png]
Figure 17
Figure 17. Figure 17: Illustration of inference results for CFT, LCAFNet, ICAFusion, EI [PITH_FULL_IMAGE:figures/full_fig_p036_17.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

145 extracted references · 73 canonical work pages

  1. [1]

    C. Bao, J. Cao, Q. Hao, Y . Cheng, Y . Ning, T. Zhao, Dual-yolo architecture from infrared and visible images for object detection, Sensors 23 (6) (2023) 2934

    2023

  2. [2]

    K. Hu, Y . He, Y . Li, J. Zhao, S. Chen, Y . Kang, Ei 2 det: Edge-guided illumination-aware interactive learning for visible-infrared object detection, IEEE Transactions on Circuits and Systems for Video Technology (2025)

    2025

  3. [3]

    Zhang, X

    J. Zhang, X. Song, Y . Li, D. Liang, Z. Zhang, J. Cai, Adaptive dual cross-attention network for multispectral object detection in autonomous driving, Expert Systems with Applications (2026) 132012

    2026

  4. [4]

    Z. Tang, Z. Wu, M. Li, J. Wen, B. Zhang, Y . Xu, J. Li, Adaptive fine-grained fusion network for multimodal uav object detection, IEEE Transactions on Image Pro- cessing (2026)

    2026

  5. [5]

    G. Li, G. Ren, J. Wang, M. Zhi, Z. Yu, B. Jiang, H. Guan, Q. Guo, Cross-modal edge-enhanced detector for uav- based multispectral object detection, Scientific Reports (2025)

    2025

  6. [6]

    Zhang, J

    Y . Zhang, J. Chen, J. Wang, D. Shi, S. Han, L. Deng, C2dff-net for object detection in multimodal remote sensing images, IEEE Transactions on Geoscience and Remote Sensing 63 (2025) 1–16.doi:10.1109/TGRS. 2025.3614295

    arXiv 2025

  7. [7]

    S. Wang, G. Sun, L. Dong, B. Zheng, Carnet: Cross- attention guided feature reconstruction for rgbt object de- tection, Available at SSRN 5292837

  8. [8]

    H. Li, L. Xiao, L. Cao, D. Wu, Y . Liu, Y . Li, Y . Zhang, H. Bao, Crossmodalnet: A dual-modal object detec- tion network based on cross-modal fusion and channel interaction, Expert Systems with Applications (2025) 129677

    2025

Show all 145 references
  1. [9]

    J. Ma, P. Hu, Dmfusion-yolov8: A difference-aware modality fusion framework for infrared–visible object detection, Signal, Image and Video Processing 19 (18) (2025) 1448. 37

    2025

  2. [10]

    Z. Chen, Y . Qian, X. Yang, C. Wang, M. Yang, Amfd: Distillation via adaptive multimodal fusion for multi- spectral pedestrian detection, IEEE Transactions on Mul- timedia (2025)

    2025

  3. [11]

    Zhang, K

    L. Zhang, K. Dong, Y . Song, G. Zhang, G. Yan, T. Liu, Y . Wang, Y . Li, X. Li, Multimodal object detection method based on bidirectional dynamic sampling and adaptive cross-modal fusion, Optics & Laser Technology 192 (2025) 113996

    2025

  4. [12]

    M. Cui, J. Nie, H. Sun, J. Xie, J. Cao, Y . Pang, X. Li, Multispectral remote sensing object detection via se- lective cross-modal interaction and aggregation, Neural Networks (2025) 108533

    2025

  5. [13]

    C. Liu, X. Ma, X. Yang, Y . Zhang, Y . Dong, Como: Cross-mamba interaction and offset-guided fusion for multimodal object detection, Information Fusion 125 (2026) 103414

    2026

  6. [14]

    H. Xu, X. Yuan, J. Wang, Y . Wang, Scvi: A semi- coupled visible-infrared small object detection method based on multimodal proposal-level probability fusion strategy, Neurocomputing (2026) 132688

    2026

  7. [15]

    J. Hu, L. Ni, B. Peng, T. Li, Wtcafnet: A wavelet trans- form and cross-attention modality-adaptive fusion net- work for multispectral object detection, Signal Process- ing (2025) 110446

    2025

  8. [16]

    E. Wang, J. Li, T. Yu, S. Xu, P. Qu, L. Na, Y . Chen, Y . Cheng, A uav detection in complex environments method based on cross-modal fusion of infrared and vis- ible images, Signal, Image and Video Processing 20 (2) (2026) 69

    2026

  9. [17]

    W. Wu, X. Zhang, H. Yin, H. Zeng, C. Wei, L. Yu, Y . Zhang, Cdfnet: Cross-dimension fusion network with dual feature enhancement for multimodal object detec- tion, Expert Systems with Applications (2026) 132380

    2026

  10. [18]

    C. Zhao, B. Mo, J. Zhao, Y . Tao, D. Zhao, Cmifdf: A lightweight cross-modal image fusion and weight- sharing object detection network framework, Infrared Physics & Technology 145 (2025) 105631

    2025

  11. [19]

    X. Duan, X. Liu, Z. Li, J. Lei, S. Li, J. Zhang, Dfas- cma: A decoupled feature adaptive sharing and cross- modulation approach for uav multi-source object detec- tion, Information Fusion (2026) 104320

    2026

  12. [20]

    X. Chen, S. Xu, S. Hu, X. Ma, Dgfd: A dual-graph con- volutional network for image fusion and low-light object detection, Information Fusion 119 (2025) 103025

    2025

  13. [21]

    Cheng, H

    M. Cheng, H. Huang, X. Liu, H. Mo, X. Zhao, S. Wu, Lefuse: Joint low-light enhancement and image fusion for nighttime infrared and visible images, Neurocomput- ing 626 (2025) 129592

    2025

  14. [22]

    Y . Shao, Q. Huang, et al., Mod-yolo: Multispectral ob- ject detection based on transformer dual-stream yolo, Pattern Recognition Letters 183 (2024) 26–34

    2024

  15. [23]

    Cheng, B

    B. Cheng, B. Xu, W. Gan, Q. Wang, Multimodal collab- orative interactive soft fusion network for rgb-infrared aerial image object detection, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sens- ing 19 (2025) 3206–3218

    2025

  16. [24]

    K. Bai, L. He, S. Ma, J. Dang, Osfgnet: Object saliency- driven feature aggregation network for multi-modal re- mote sensing object detection, IEEE Geoscience and Re- mote Sensing Letters (2026)

    2026

  17. [25]

    X. Wang, J. Xi, F. Yang, Y . Yang, M. Li, Pfi-net: A par- allel feature interaction network for infrared and visible target detection, Pattern Recognition (2025) 113003

    2025

  18. [26]

    M. Yuan, X. Shi, N. Wang, Y . Wang, X. Wei, Improving rgb-infrared object detection with cascade alignment- guided transformer, Information Fusion 105 (2024) 102246

    2024

  19. [27]

    F. Yang, W. Li, L. Li, M. Yang, J. Zhang, Dwsf-net: A dynamic wavelet-based spatial-frequency fusion net- work for multispectral object detection, IEEE Transac- tions on Multimedia (2026)

    2026

  20. [28]

    Z. Wang, T. Tian, Regional defeats global: An effi- cient regional feature fusion via convolutional architec- ture for multispectral object detection, Information Fu- sion (2026) 104110

    2026

  21. [29]

    X. Wu, L. Wang, J. Guan, H. Ji, L. Xu, Y . Hou, A. Fei, Dhanet: Dual-stream hierarchical interaction networks for multimodal drone object detection, IEEE Transac- tions on Geoscience and Remote Sensing (2025)

    2025

  22. [30]

    Zhang, M

    J. Zhang, M. Gao, Y . Wang, C. Li, D. Fang, Z. Wei, Hffn: Hierarchical feature fusion network for rgb-infrared ob- ject detection, Infrared Physics & Technology (2026) 106429

    2026

  23. [31]

    H. Wu, P. Yuan, W. Liu, A. Wang, L. Yu, G. Molnar, Ivd- net: An adaptive dual-branch network for small object detection in multi-modal remote sensing images, IEEE Journal of Selected Topics in Applied Earth Observa- tions and Remote Sensing (2026)

    2026

  24. [32]

    Z. Tang, Y . Xie, T. Xu, X.-J. Wu, J. Kittler, Learning bi-directional fusion and deformation-sensitive loss for rgb-t tiny object detection, Information Fusion (2025) 103985

    2025

  25. [33]

    W. Wu, H. Zhang, X. Zhang, H. Yin, Y . Zhang, Lightweight modal-guided cross-attention fusion net- work for visible-infrared object detection, Pattern Recognition (2026) 113350. 38

    2026

  26. [34]

    J. Liu, J. Ni, Z. Zhang, Y . Gu, S. X. Yang, Maftnet: Mul- timodal adaptive fusion-based transformer network for infrared and visible image uav object detection, IEEE Sensors Journal (2026)

    2026

  27. [35]

    Z. Hou, J. Zhao, X. Li, S. Ma, X. Yang, L. Pu, Mul- timodal detection transformer with multiscale cross- modal feature fusion and selective query recollection, In- frared Physics & Technology (2025) 106232

    2025

  28. [36]

    Wang, I.-H

    C.-Y . Wang, I.-H. Yeh, H.-Y . Mark Liao, Yolov9: Learn- ing what you want to learn using programmable gradient information, in: European conference on computer vi- sion, Springer, 2024, pp. 1–21

    2024

  29. [37]

    Wang, H.-Y

    C.-Y . Wang, H.-Y . M. Liao, Y .-H. Wu, P.-Y . Chen, J.-W. Hsieh, I.-H. Yeh, Cspnet: A new backbone that can en- hance learning capability of cnn, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops, 2020, pp. 390–391

    2020

  30. [38]

    C.-Y . Wang, A. Bochkovskiy, H.-Y . M. Liao, Yolov7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023, pp. 7464–7475

    2023

  31. [39]

    X. Ding, X. Zhang, N. Ma, J. Han, G. Ding, J. Sun, Repvgg: Making vgg-style convnets great again, in: Pro- ceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, 2021, pp. 13733–13742

    2021

  32. [40]

    A. Wang, H. Chen, Z. Lin, J. Han, G. Ding, Repvit: Re- visiting mobile cnn from vit perspective, in: Proceed- ings of the IEEE/CVF conference on computer vision and pattern recognition, 2024, pp. 15909–15920

    2024

  33. [41]

    J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, L. Fei-Fei, Imagenet: A large-scale hierarchical image database, in: 2009 IEEE conference on computer vision and pattern recognition, Ieee, 2009, pp. 248–255

    2009

  34. [42]

    J. Hu, L. Shen, G. Sun, Squeeze-and-excitation net- works, in: Proceedings of the IEEE conference on com- puter vision and pattern recognition, 2018, pp. 7132– 7141

    2018

  35. [43]

    Vaswani, N

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, I. Polosukhin, At- tention is all you need, Advances in neural information processing systems 30 (2017)

    2017

  36. [44]

    W. Luo, Y . Li, R. Urtasun, R. Zemel, Understanding the effective receptive field in deep convolutional neural net- works, Advances in neural information processing sys- tems 29 (2016)

    2016

  37. [45]

    F. Yu, V . Koltun, Multi-scale context aggregation by dilated convolutions, arXiv preprint arXiv:1511.07122 (2015)

    2015 arXiv

  38. [46]

    Zhang, X

    X. Zhang, X. Zhou, M. Lin, J. Sun, Shufflenet: An ex- tremely efficient convolutional neural network for mo- bile devices, in: Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 6848–6856

    2018

  39. [47]

    Touvron, M

    H. Touvron, M. Cord, A. Sablayrolles, G. Synnaeve, H. Jégou, Going deeper with image transformers, in: Proceedings of the IEEE/CVF international conference on computer vision, 2021, pp. 32–42

    2021

  40. [48]

    Y . Sun, B. Cao, P. Zhu, Q. Hu, Drone-based rgb-infrared cross-modality vehicle detection via uncertainty-aware learning, IEEE Transactions on Circuits and Systems for Video Technology 32 (10) (2022) 6700–6713

    2022

  41. [49]

    Ouyang, S

    D. Ouyang, S. He, G. Zhang, M. Luo, H. Guo, J. Zhan, Z. Huang, Efficient multi-scale attention module with cross-spatial learning, in: ICASSP 2023-2023 IEEE in- ternational conference on acoustics, speech and signal processing (ICASSP), IEEE, 2023, pp. 1–5

    2023

  42. [50]

    W. Xu, Y . Wan, W. Zhao, Ela: efficient location attention for deep convolution neural networks, Journal of Real- Time Image Processing 22 (4) (2025) 1–14

    2025

  43. [51]

    Q. Hou, D. Zhou, J. Feng, Coordinate attention for ef- ficient mobile network design, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 13713–13722

    2021

  44. [52]

    Abdelfattah, A

    A. Abdelfattah, A. Haidar, S. Tomov, J. Dongarra, Per- formance, design, and autotuning of batched gemm for gpus, in: International Conference on High Performance Computing, Springer, 2016, pp. 21–38

    2016

  45. [53]

    Huang, Z

    G. Huang, Z. Liu, L. Van Der Maaten, K. Q. Weinberger, Densely connected convolutional networks, in: Proceed- ings of the IEEE conference on computer vision and pat- tern recognition, 2017, pp. 4700–4708

    2017

  46. [54]

    S. Woo, J. Park, J.-Y . Lee, I. S. Kweon, Cbam: Convo- lutional block attention module, in: Proceedings of the European conference on computer vision (ECCV), 2018, pp. 3–19

    2018

  47. [55]

    C. E. Shannon, A mathematical theory of communica- tion, The Bell system technical journal 27 (3) (1948) 379–423

    1948

  48. [56]

    Kullback, Information theory and statistics, Courier Corporation, 1997

    S. Kullback, Information theory and statistics, Courier Corporation, 1997

    1997

  49. [57]

    Z. Wang, A. C. Bovik, H. R. Sheikh, E. P. Simoncelli, Image quality assessment: from error visibility to struc- tural similarity, IEEE transactions on image processing 13 (4) (2004) 600–612

    2004

  50. [58]

    M. A. K. Raiaan, S. Sakib, N. M. Fahad, A. Al Mamun, M. A. Rahman, S. Shatabda, M. S. H. Mukta, A system- atic review of hyperparameter optimization techniques in 39 convolutional neural networks, Decision Analytics Jour- nal 11 (2024) 100470

    2024

  51. [59]

    Zheng, P

    Z. Zheng, P. Wang, D. Ren, W. Liu, R. Ye, Q. Hu, W. Zuo, Enhancing geometric factors in model learning and inference for object detection and instance segmen- tation, IEEE transactions on cybernetics 52 (8) (2021) 8574–8586

    2021

  52. [60]

    X. Li, W. Wang, L. Wu, S. Chen, X. Hu, J. Li, J. Tang, J. Yang, Generalized focal loss: Learning qualified and distributed bounding boxes for dense object detection, Advances in neural information processing systems 33 (2020) 21002–21012

    2020

  53. [61]

    Teledyne FLIR, FREE Teledyne FLIR Thermal Dataset for Algorithm Training, Available online: https://oem.flir.com/solutions/automotive/ adas-dataset-form/, (Accessed on 25 December

  54. [62]

    J. Liu, X. Fan, Z. Huang, G. Wu, R. Liu, W. Zhong, Z. Luo, Target-aware dual adversarial learning and a multi-scenario multi-modality benchmark to fuse in- frared and visible for object detection, in: Proceedings of the IEEE/CVF conference on computer vision and pat- tern reco...

    2022

  55. [63]

    X. Jia, C. Zhu, M. Li, W. Tang, W. Zhou, Llvip: A visible-infrared paired dataset for low-light vision, in: Proceedings of the IEEE/CVF international conference on computer vision, 2021, pp. 3496–3504

    2021

  56. [64]

    Razakarivony, F

    S. Razakarivony, F. Jurie, Vehicle detection in aerial imagery: A small target detection benchmark, Journal of Visual Communication and Image Representation 34 (2016) 187–203

    2016

  57. [65]

    Zhang, E

    H. Zhang, E. Fromont, S. Lefevre, B. Avignon, Multi- spectral fusion for object detection with cyclic fuse-and- refine blocks, in: 2020 IEEE International conference on image processing (ICIP), IEEE, 2020, pp. 276–280

    2020

  58. [66]

    Everingham, L

    M. Everingham, L. Van Gool, C. K. Williams, J. Winn, A. Zisserman, The pascal visual object classes (voc) challenge, International journal of computer vision 88 (2) (2010) 303–338

    2010

  59. [67]

    Girshick, Fast r-cnn, in: Proceedings of the IEEE international conference on computer vision, 2015, pp

    R. Girshick, Fast r-cnn, in: Proceedings of the IEEE international conference on computer vision, 2015, pp. 1440–1448

    2015

  60. [68]

    W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. Reed, C.- Y . Fu, A. C. Berg, Ssd: Single shot multibox detector, in: European conference on computer vision, Springer, 2016, pp. 21–37

    2016

  61. [69]

    T.-Y . Lin, P. Goyal, R. Girshick, K. He, P. Dollár, Focal loss for dense object detection, in: Proceedings of the IEEE international conference on computer vision, 2017, pp. 2980–2988

    2017

  62. [70]

    Jocher, A

    G. Jocher, A. Chaurasia, A. Stoken, J. Borovec, Y . Kwon, K. Michael, J. Fang, C. Wong, Z. Yifu, D. Montes, et al., ultralytics/yolov5: v6. 2-yolov5 clas- sification models, apple m1, reproducibility, clearml and deci. ai integrations, Zenodo (2022)

    2022

  63. [71]

    Jocher, A

    G. Jocher, A. Chaurasia, J. Qiu, Ultralytics yolo,https: //github.com/ultralytics/ultralytics(2023)

    2023

  64. [72]

    Khanam, M

    R. Khanam, M. Hussain, Yolov11: An overview of the key architectural enhancements, arXiv preprint arXiv:2410.17725 (2024)

    2024 arXiv

  65. [73]

    Z. Ge, S. Liu, F. Wang, Z. Li, J. Sun, Yolox: Exceed- ing yolo series in 2021, arXiv preprint arXiv:2107.08430 (2021)

    2021 arXiv

  66. [74]

    Zhang, F

    H. Zhang, F. Li, S. Liu, L. Zhang, H. Su, J. Zhu, L. M. Ni, H.-Y . Shum, Dino: Detr with improved denoising anchor boxes for end-to-end object detection, arXiv preprint arXiv:2203.03605 (2022)

    2022 arXiv

  67. [75]

    Y . Chen, X. Dai, D. Chen, M. Liu, X. Dong, L. Yuan, Z. Liu, Mobile-former: Bridging mobilenet and trans- former, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 5270–5279

    2022

  68. [76]

    X. Liu, H. Peng, N. Zheng, Y . Yang, H. Hu, Y . Yuan, Ef- ficientvit: Memory efficient vision transformer with cas- caded group attention, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023, pp. 14420–14430

    2023

  69. [77]

    Zhang, J

    J. Zhang, J. Lei, W. Xie, Z. Fang, Y . Li, Q. Du, Supery- olo: Super resolution assisted object detection in mul- timodal remote sensing imagery, IEEE Transactions on Geoscience and Remote Sensing 61 (2023) 1–15

    2023

  70. [78]

    Y . Zeng, T. Liang, Y . Jin, Y . Li, Mmi-det: Exploring multi-modal integration for visible and infrared object detection, IEEE Transactions on Circuits and Systems for Video Technology 34 (11) (2024) 11198–11213

    2024

  71. [79]

    Qingyun, H

    F. Qingyun, H. Dapeng, W. Zhaokui, Cross-modality fu- sion transformer for multispectral object detection, arXiv preprint arXiv:2111.00273 (2021)

    2021 arXiv

  72. [80]

    J. Shen, Y . Chen, Y . Liu, X. Zuo, H. Fan, W. Yang, Ica- fusion: Iterative cross-attention guided feature fusion for multispectral object detection, Pattern Recognition 145 (2024) 109913

    2024

  73. [81]

    Y . Cao, J. Bin, J. Hamari, E. Blasch, Z. Liu, Multi- modal object detection by channel switching and spatial attention, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 403–411. 40

    2023

  74. [82]

    H. Fu, S. Wang, P. Duan, C. Xiao, R. Dian, S. Li, Z. Li, Lraf-net: Long-range attention fusion network for visible–infrared object detection, IEEE Transactions on Neural Networks and Learning Systems 35 (10) (2023) 13232–13245

    2023

  75. [83]

    H. Fu, H. Liu, J. Yuan, X. He, J. Lin, Z. Li, Yolo- adaptor: a fast adaptive one-stage detector for non- aligned visible-infrared object detection, IEEE Transac- tions on Intelligent Vehicles (2024)

    2024

  76. [84]

    S. Lee, J. Park, J. Park, Crossformer: Cross-guided at- tention for multi-modal object detection, Pattern Recog- nition Letters 179 (2024) 144–150

    2024

  77. [85]

    J. Jang, J. Lee, J. Paik, Camdet: Condition-adaptive mul- tispectral object detection using a visible-thermal trans- lation model, in: ICASSP 2025-2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, 2025, pp. 1–5

    2025

  78. [86]

    C. Sun, Y . Chen, X. Qiu, R. Li, L. You, Mrd-yolo: A multispectral object detection algorithm for complex road scenes, Sensors 24 (10) (2024) 3222

    2024

  79. [87]

    Y . Zhao, Y . Gao, X. Yang, L. Yang, Multispectral target detection based on deep feature fusion of visible and in- frared modalities, Applied Sciences 15 (11) (2025) 5857

    2025

  80. [88]

    X. He, T. Yang, T. Yan, H. Li, Y . Ge, Z. Ren, Z. Liu, J. Jiang, C. Tang, Efficient layer-wise cross-view calibra- tion and aggregation for multispectral object detection, Electronics 15 (3) (2026) 498

    2026

  81. [89]

    X. Yang, Y . Liu, W. Pan, G. Chu, J. Zhang, J. Zhao, Z. Man, X. Cao, M2i2ha: A multi-modal object detec- tion method based on intra-and inter-modal hypergraph attention, arXiv preprint arXiv:2601.14776 (2026)

    2026 arXiv

  82. [90]

    W. Xu, Y . Yang, Jfdet: Joint fusion and detection for multimodal remote sensing imagery, Remote Sensing 18 (1) (2026) 176

    2026

  83. [91]

    J. Jang, C. Park, H. Kim, J. Lee, J. Paik, Multispec- tral object detection enhanced by cross-modal informa- tion complementary and cosine similarity channel re- sampling modules, in: 2025 IEEE/CVF Winter Con- ference on Applications of Computer Vision (W ACV), IEEE, 2025, pp....

    2025

  84. [92]

    Zhang, L

    Q. Zhang, L. Tabaro, A. M. Abdelmoniem, J. An, Dlrmamba: Distilling low-rank mamba for edge multispectral fusion object detection, arXiv preprint arXiv:2603.06920 (2026)

    2026

  85. [93]

    K. Li, Z. Zhong, Z. Luo, H. Tian, K. Wang, H. Jiang, D. Xiang, W. Tang, Fcat: Frequency-domain cross- attention for all-weather multispectral object detection in low-altitude uav security inspection of urban and indus- trial areas, Remote Sensing 18 (5) (2026) 826

    2026

  86. [94]

    J. Li, C. Sui, J. Wang, J. Zhou, Pmdet: Patch-aware en- hancement and fusion for multispectral object detection, Remote Sensing 18 (7) (2026) 1068

    2026

  87. [95]

    Y . Chen, J. Ye, X. Wan, Tf-yolo: a transformer–fusion- based yolo detector for multimodal pedestrian detection in autonomous driving scenes, World Electric Vehicle Journal 14 (12) (2023) 352

    2023

  88. [96]

    Zhang, H

    Y . Zhang, H. Yu, Y . He, X. Wang, W. Yang, Illumination-guided rgbt object detection with inter-and intra-modality fusion, IEEE Transactions on Instrumen- tation and Measurement 72 (2023) 1–13

    2023

  89. [97]

    F. Yang, B. Liang, W. Li, J. Zhang, Multidimensional fusion network for multispectral object detection, IEEE Transactions on Circuits and Systems for Video Technol- ogy (2024)

    2024

  90. [98]

    H. Zhu, Y . Chen, X. Pan, Y . He, J. Wu, Modality-guided feature alignment and complementary enhancement for infrared-visible object detection, IEEE Transactions on Circuits and Systems for Video Technology (2026)

    2026

  91. [99]

    F. Liu, C. Gao, F. Chen, P. Li, J. Guo, D. Meng, A fusion- enhanced network for infrared and visible high-level vi- sion tasks, IEEE Transactions on Multimedia (2025)

    2025

  92. [100]

    K. Li, D. Wang, Z. Hu, S. Li, W. Ni, L. Zhao, Q. Wang, Fd2-net: Frequency-driven feature decomposition net- work for infrared-visible object detection, in: Proceed- ings of the AAAI Conference on Artificial Intelligence, V ol. 39, 2025, pp. 4797–4805

    2025

  93. [101]

    W. Wu, X. Zhang, H. Yin, S. Dai, H. Zhang, Y . Zhang, Fredft: Frequency domain fusion transformer for visible-infrared object detection, arXiv preprint arXiv:2511.10046 (2025)

    2025

  94. [102]

    H. Yang, J. Fang, Y . Zhu, X. Zhao, Y . Guo, X. Zhang, X. Hu, X. Yang, Q. Ming, Crossweaver: Towards efficient cross-modal interweaving and decoupling for weakly-aligned multispectral object detection, in: Pro- ceedings of the IEEE/CVF Conference on Computer Vi- sion and Patte...

    2026

  95. [103]

    S. Peng, R. Xue, Y . Tong, Z. Wang, H. Yang, Multispec- tral object detection via edge-enhanced and frequency- aware fusion network, IEEE Journal of Selected Top- ics in Applied Earth Observations and Remote Sensing (2025)

    2025

  96. [104]

    Y . Shao, T. Shi, Representation space constrained learn- ing with modality decoupling for multimodal object de- tection, arXiv preprint arXiv:2511.15433 (2025)

    2025

  97. [105]

    H. Xiao, J. Zhuang, B. Yang, J. Hu, J. Zhu, Z. Lian, S. Zhao, Y . Zhou, Saff: A spatially-aware fusion frame- work for effective and efficient aerial object detection, IEEE Transactions on Geoscience and Remote Sensing (2026). 41

    2026

  98. [106]

    Huang, Z

    K. Huang, Z. Zhang, C. Shi, L. Luo, J. Shi, Y . Liu, De- tection drives an end-to-end fusion of infrared and visible images based on diffusion models, IEEE Transactions on Image Processing (2026)

    2026

  99. [107]

    L. Lu, S. Zhang, Y . Gu, B. Liu, B. Liu, A cross-modal hierarchical enhanced fusion method for object detection in intelligent transportation systems, Digital Signal Pro- cessing (2026) 106094

    2026

  100. [108]

    C. Yuan, W. Li, Q. Zhou, Q. Li, C. Li, T. Gong, A fusion-enhanced infrared-visible object detection net- work with target difference sensitivity, Neurocomputing (2026) 133640

    2026

  101. [109]

    Z. Zhu, X. Song, G. Zhou, A. Gong, Q. Zhu, Acse: Ad- vantage complementation and salient feature-enhanced fusion for visible-infrared object detection, Journal of King Saud University Computer and Information Sci- ences (2026)

    2026

  102. [110]

    Y . Hu, H. Jin, Modality-aware fusion and selection for robust multispectral pedestrian detection (2026)

    2026

  103. [111]

    H. Wang, X. Liu, Mutual distillation attribute fusion network for multimodal vehicle object detection, IEEE Transactions on Vehicular Technology (2026)

    2026

  104. [112]

    X. Wang, X. Chen, H. Fan, W. Ren, S. Wang, Y . Tang, L. Liu, Z. Han, Seeing only the focus: Rgb-t object- aware region enhancement for object detection in harsh environments, IEEE Transactions on Multimedia (2026)

    2026

  105. [113]

    Y . Liu, G. Sun, Q. Guo, Y . Wu, Df-net: A dual-modal de- tection network based on decoupled deformable dilated convolution and progressive coordinated fusion, IEEE Sensors Journal (2026)

    2026

  106. [114]

    C. Yang, X. Zhang, Y . Xiao, F. Meng, Wd-fqdet: Mul- tispectral detection transformer via wavelet decomposi- tion and frequency-aware query learning, arXiv preprint arXiv:2605.13621 (2026)

    2026 arXiv

  107. [115]

    P. Lu, S. Yu, Z. Li, Alignfree-net: A registration-free hybrid framework for infrared-visible object detection via semantic feature matching, in: 2026 11th Interna- tional Conference on Intelligent Computing and Signal Processing (ICSP), IEEE, 2026, pp. 2107–2111

    2026

  108. [116]

    H. R. Medeiros, F. A. G. Pena, M. Aminbeidokhti, T. Dubail, E. Granger, M. Pedersoli, Hallucidet: hallu- cinating rgb modality for person detection through priv- ileged information, in: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2024, pp....

    2024

  109. [117]

    Z. Wang, F. Colonnier, J. Zheng, J. Acharya, W. Jiang, K. Huang, Tirdet: Mono-modality thermal infrared ob- ject detection based on prior thermal-to-visible transla- tion, in: Proceedings of the 31st ACM International Con- ference on Multimedia, 2023, pp. 2663–2672

    2023

  110. [118]

    Z. Wang, H. Shen, W. Jiang, K. Huang, A fourier- transform-based framework with asymptotic attention for mobile thermal infrared object detection, IEEE Sen- sors Journal 24 (13) (2024) 21012–21024

    2024

  111. [119]

    H. Liu, F. Jin, H. Zeng, H. Pu, B. Fan, Image en- hancement guided object detection in visually degraded scenes, IEEE transactions on neural networks and learn- ing systems 35 (10) (2023) 14164–14177

    2023

  112. [120]

    L. Tang, Z. Chen, J. Huang, J. Ma, Camf: An inter- pretable infrared and visible image fusion network based on class activation mapping, IEEE Transactions on Mul- timedia 26 (2023) 4776–4791

    2023

  113. [121]

    W. Zhao, S. Xie, F. Zhao, Y . He, H. Lu, Metafusion: Infrared and visible image fusion via meta-feature em- bedding from object detection, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 13955–13965

    2023

  114. [122]

    Z. Zhao, H. Bai, Y . Zhu, J. Zhang, S. Xu, Y . Zhang, K. Zhang, D. Meng, R. Timofte, L. Van Gool, Ddfm: De- noising diffusion model for multi-modality image fusion, in: Proceedings of the IEEE/CVF international confer- ence on computer vision, 2023, pp. 8082–8093

    2023

  115. [123]

    W. Dong, H. Zhu, S. Lin, X. Luo, Y . Shen, G. Guo, B. Zhang, Fusion-mamba for cross-modality object de- tection, IEEE Transactions on Multimedia (2025)

    2025

  116. [124]

    Huang, W

    Z. Huang, W. Li, Y . Zhang, J. Guo, J. Zheng, G. Ji, Y . Tao, Cdfit: A transformer using cross-modal dual- stream feature interaction for multispectral pedestrian detection, IEEE Transactions on Intelligent Transporta- tion Systems (2026)

    2026

  117. [125]

    F. Meng, A. Hong, H. Tang, G. Tong, Fqdnet: A fusion- enhanced quad-head network for rgb-infrared object de- tection, Remote Sensing 17 (6) (2025) 1095

    2025

  118. [126]

    H. Wang, L. Jin, G. Wang, W. Liu, Q. Shi, Y . Hou, J. Liu, Rgb-fir multimodal pedestrian detection with cross- modality context attentional model, Sensors 25 (13) (2025) 3854

    2025

  119. [127]

    J. Gong, Z. Yuan, W. Li, W. Li, Y . Guo, B. Guo, A lightweight upsampling and cross-modal feature fusion- based algorithm for small-object detection in uav im- agery, Electronics 15 (2) (2026) 298

    2026

  120. [128]

    C. Chai, Y . Wang, X. Hu, S. Han, Efficient multi-spectral pedestrian detection network for occlusion and far dis- tance targets, Available at SSRN 5312893

  121. [129]

    R. Lan, Y . Zhang, Z. Wang, W. Liu, R. Yang, Kdet-hpfl: A personalized federated learning framework for multi- modal pedestrian detection with adaptive feature selec- tion, IEEE Internet of Things Journal (2025). 42

    2025

  122. [130]

    T. Weng, X. Niu, Lmdenet: A lightweight rgb-ir object detection network for low-light remote sensing images, Sensors 26 (4) (2026) 1130

    2026

  123. [131]

    X. Zhai, F. Xu, Y . Pan, Mcff-det: Multispectral coarse- to-fine fusion for object detection, in: 2025 IEEE 35th International Workshop on Machine Learning for Signal Processing (MLSP), IEEE, 2025, pp. 1–6

    2025

  124. [132]

    R. Chen, H. Sun, H. Fan, P. Wu, Vif-yolo: Visible- infrared fusion yolo for remote sensing small object de- tection, in: 2025 IEEE 31th International Conference on Parallel and Distributed Systems (ICPADS), IEEE, 2025, pp. 1–8

    2025

  125. [133]

    X. Guo, F. Yang, L. Ji, Def-net: A dual-modal feature enhancement and fusion network for infrared and visible object detection, PloS one 21 (4) (2026) e0345815

    2026

  126. [134]

    H. Li, J. Ma, J. Zhang, H. Luo, H. Zuo, Y . Wu, T. Tan, Hyperdet: Cross-modal hypergraph fusion for enhanced object detection in uav imagery, Journal of King Saud University Computer and Information Sciences (2026)

    2026

  127. [135]

    Q. Hu, H. Yu, Z. Zhou, S. Li, Iaf-rtdetr: Illumina- tion evaluation-driven multimodal object detection net- work for infrared–visible dual-source fusion, Electronics 15 (6) (2026) 1332

    2026

  128. [136]

    Zhang, G

    Z. Zhang, G. Lv, Y . Gao, G. Ma, Mdsf-det: Modality decoupling and synergistic fusion detector, in: ICASSP 2026-2026 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, 2026, pp. 10377–10381

    2026

  129. [137]

    C. Guo, T. Zhao, Z. Zhang, Yolo-msff: An improved yolo11-based multispectral feature fusion method for object detection, in: 2026 International Conference on Generative Artificial Intelligence and Information Secu- rity (GAIIS), IEEE, 2026, pp. 574–583

    2026

  130. [138]

    H. Wu, J. Rong, Q. Zhou, D. Yang, W. Wang, A guided fusion network based on cross-scale semantic alignment for multi-spectral object detection, Pattern Recognition (2026) 114024

    2026

  131. [139]

    X. Xue, H. Chen, K. Song, Y . Yan, B. Li, Q. Meng, Transfer graph reasoning network for misaligned visible- thermal object detection, Knowledge-Based Systems (2026) 116271

    2026

  132. [140]

    L. Yu, Z. Han, Edge-guided progressive feature comple- mentary network for visible–infrared remote sensing ob- ject detection, International Journal of Remote Sensing (2026) 1–28

    2026

  133. [141]

    H. Guo, C. Sun, J. Zhang, W. Zhang, N. Zhang, Mmyfnet: Multi-modality yolo fusion network for ob- ject detection in remote sensing images, Remote Sensing 16 (23) (2024) 4451

    2024

  134. [142]

    Z. Wu, F. Zhang, B. Yang, Multimodal uav object de- tection based on progressive bi-directional attention and relational bayesian constraints (2026)

    2026

  135. [143]

    Cheng, Y

    Q. Cheng, Y . Jiang, Y . Gao, Y . Qiu, Y . Tang, X. Tu, Yolo- ch: A cross-modal feature interaction and screening- based dual-stream network for uav small object detec- tion, Drones 10 (5) (2026) 350

    2026

  136. [144]

    Jocher, A

    G. Jocher, A. Chaurasia, A. Stoken, J. Borovec, Y . Kwon, K. Michael, J. Fang, Z. Yifu, C. Wong, D. Montes, et al., ultralytics/yolov5: v7. 0-yolov5 sota realtime instance segmentation, Zenodo (2022)

    2022

  137. [145]

    A. Wang, H. Chen, L. Liu, K. Chen, Z. Lin, J. Han, G. Ding, Yolov10: Real-time end-to-end object detec- tion, Advances in neural information processing systems 37 (2024) 107984–108011. Nima Hatamireceived the B.Sc. de- gree in Biomedical Engineering and the M.Sc. degree in Ele...

    2024

Pith tools

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