LDFE: Laplacian Decoupled Feature Enhancement Block for Dual-Stream CNN-based RGB-IR Object Detection

Baochang Zhang; Guodong Guo; Haodong Zhu; Linlin Yang; Wenhao Dong; Xiaorong Shi; Xiaoyan Luo

REVIEW 3 major objections 3 minor 54 references

LDFE: Laplacian Decoupled Feature Enhancement Block for Dual-Stream CNN-based RGB-IR Object Detection

T0 review · 3 major / 3 minor · reviewed 2026-07-10 · glm-5.2

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

Pith's one-line read Laplacian decomposition lifts RGB-IR detection up to 6.2%

desk verdict Architectural combination of Laplacian decomposition with state-space models for RGB-IR fusion is new; empirical claims are unverifiable from abstract alone. read the letter →

arxiv 2607.08076 v1 pith:DZVYWNN6 submitted 2026-07-09 cs.CV cs.AI

Wenhao Dong , Xiaoyan Luo , Linlin Yang , Haodong Zhu , Xiaorong Shi , Guodong Guo , Baochang Zhang This is my paper · ORCID

classification cs.CVcs.AI

keywords featurefusionenhancementldfemainauxiliarydual-streamfeatures

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

The paper proposes the Laplacian Decoupled Feature Enhancement (LDFE) block, a plug-in module for dual-stream CNN object detectors that fuse visible-light (RGB) and infrared (IR) imagery. The core idea is to stop treating feature fusion as a single monolithic operation. Instead, LDFE uses a Laplacian Pyramid to split each modality's feature maps into a global (low-frequency) component and a local (high-frequency) component, then applies different fusion strategies to each. Global features are handled by a Global State Space Enhancement module (GS2E): one modality acts as the main signal and the other as auxiliary, with the auxiliary generating cross-modal attention that suppresses noise in the main while a State Space Model captures long-range dependencies. The two modalities alternate main/auxiliary roles so the interaction is bidirectional. Local features are handled by a Local Convolutional Correlation Enhancement module (LC2E), which denoises fine-grained spatial detail and fuses it through channel and spatial attention plus triple convolution. The decomposed features are then reconstructed. The authors report mAP improvements of 6.2%, 3.7%, 4.7%, 2.3%, 4.1%, and 2.0% over prior state-of-the-art methods on six RGB-IR detection benchmarks (M3FD, DroneVehicle, LLVIP, FLIR-Aligned, KAIST, and VEDAI).

What carries the argument

LDFE block (Laplacian Pyramid decomposition → GS2E for global features with bidirectional cross-modal attention + State Space Model → LC2E for local features with triple convolution → reconstruction)

What would settle it

If the Laplacian global-local decomposition is replaced by a single-band fusion pathway with equivalent parameter count, the mAP gains should largely disappear — or if they persist, the decomposition is not the operative mechanism.

Watch

Extended reading notes

Core claim

The paper's central mechanism is the Laplacian Pyramid decomposition of fused multi-modal features into global and local frequency bands, followed by band-specific denoising and fusion. The claim is that RGB and IR modalities carry different kinds of useful information at different spatial scales — global context (where objects are in the scene) versus local detail (what fine-grained edges and textures distinguish them) — and that forcing both through the same fusion pathway wastes one or both. By routing global features through a State Space Model with cross-modal attention and local features through convolutional correlation, and by making the two modalities alternate as noise-suppressor /

Load-bearing premise

The reported mAP gains over prior methods are large (up to 6.2 points), and the comparison's fairness depends on whether competing methods were retrained under identical training schedules, data augmentation, and backbone capacities — details not verifiable from the abstract alone.

Editorial extensions

If this is right

The Laplacian decomposition strategy could generalize beyond RGB-IR to other multi-modal fusion tasks (RGB-depth, RGB-event, audio-visual) where modalities contribute differently at different spatial scales.
The use of State Space Models for global feature modeling in detection pipelines offers a lower-complexity alternative to self-attention for capturing long-range dependencies in feature maps.
The bidirectional main/auxiliary role alternation between modalities suggests that asymmetric noise suppression is more effective when neither modality is permanently privileged, which could inform fusion design in sensor-swap or degradation scenarios.

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.

Desk Editor's Note

read the letter

Here's the situation: this paper proposes a Laplacian Decoupled Feature Enhancement (LDFE) block for RGB-IR object detection that combines Laplacian Pyramid decomposition with a state-space model (Mamba-style) for cross-modal feature denoising and fusion. The specific combination — splitting features into global/local components via Laplacian decomposition, then using a state-space model for global denoising and triple convolutions for local enhancement — is not something I've seen in the prior RGB-IR fusion literature. That's the real contribution here, and it's a reasonable architectural idea worth taking seriously. The design is internally coherent: global features get long-range dependency modeling through the SSM, local features get fine-grained convolutional enhancement, and the two modalities alternate main/auxiliary roles for bidirectional interaction. That's a clean design story. The claimed mAP improvements across six datasets (2.0–6.2% over SOTA) are substantial if real. Six benchmarks is a lot of ground to cover, and the breadth itself is worth credit — assuming the comparisons are honest. Now the soft spots. The reader's stress-test concern about fair baselines is the right one to flag, and it's load-bearing. In YOLO-based detection, mAP shifts of 2–6 points can easily come from training schedule differences, input resolution, backbone width, or augmentation alone — not from the architectural module. The abstract gives zero information about whether SOTA baselines were retrained under matched settings or whether numbers were pulled from original papers with different configs. Without the full text, tables, ablations, and ideally code, the empirical case is unverified. I also note that the module introduces several free parameters (Laplacian levels, SSM dimension, kernel sizes, attention parameters) with no indication of how they were selected. That's a secondary concern — typical for this kind of work — but it matters for reproducibility. The reader rated soundness at 3.0, which I think is fair given abstract-only access. The low confidence is appropriate. I don't see any architectural red flags or internal inconsistencies — the concern is purely empirical controls. This paper is for researchers working on multi-modal detection, particularly RGB-IR fusion for autonomous driving and surveillance. A reader looking for deployable architectural improvements would get value from the design ideas even if the empirical margins shrink under tighter controls. It deserves a serious referee who can check the full experimental setup, ablations, and baseline fairness. I'd recommend sending it out for full review.

Referee Report

3 major / 3 minor

Summary. The manuscript proposes a Laplacian Decoupled Feature Enhancement (LDFE) block for dual-stream CNN-based RGB-IR object detection. The LDFE block decomposes features into global and local components via a Laplacian Pyramid, then applies a Global State Space Enhancement module (GS2E) using a State Space Model for long-range dependencies and cross-modal attention, and a Local Convolutional Correlation Enhancement module (LC2E) for fine-grained detail extraction. The two modalities alternate main/auxiliary roles for bidirectional interaction. The authors report mAP improvements of 6.2%, 3.7%, 4.7%, 2.3%, 4.1%, and 2.0% over SOTA on six datasets (M3FD, DroneVehicle, LLVIP, FLIR-Aligned, KAIST, VEDAI). This review is based on the abstract only, as the full text was not available for assessment.

Significance. RGB-IR fusion for object detection is an active and practically important area. The architectural design of LDFE—combining Laplacian decomposition with modality-specific global/local enhancement and bidirectional role alternation—is a reasonable and potentially novel contribution to the feature fusion literature. The reported improvements across six datasets, if verified under controlled comparisons, would be significant. However, the significance of the contribution cannot be fully assessed without access to the full text, which is needed to verify experimental controls, ablation studies, and reproducibility. No machine-checked proofs, reproducible code, or parameter-free derivations are mentioned in the abstract.

major comments (3)

The central empirical claim rests on six mAP improvements (2.0%–6.2%) over SOTA methods. Without the full text, we cannot verify whether baselines were retrained under identical training schedules, backbone capacities, input resolutions, and data augmentation as LDFE. These confounders are known to shift mAP by several points in YOLO-based detection. The full text must explicitly state the training recipe for every baseline and confirm that SOTA numbers are reproduced from scratch under matched settings, not cited from original papers that may use different configurations. This is the single most load-bearing assumption for the paper's claims.
The abstract does not specify whether the reported improvements are absolute or relative mAP gains, nor whether error bars or statistical significance tests are reported. For a paper whose entire contribution is validated through these six numbers, the full text must include variance estimates across multiple runs and clarify whether gains are absolute or relative. Without this, the practical significance of improvements in the 2.0%–2.3% range (FLIR-Aligned, VEDAI) is difficult to assess.
No ablation study is referenced in the abstract. For a method with multiple components (Laplacian decomposition, GS2E with State Space Model, LC2E with triple convolution, bidirectional role alternation), ablation isolating each component's contribution is essential to justify the architectural complexity. The full text must include these ablations, and each component must be shown to provide a non-trivial marginal improvement.

minor comments (3)

The abstract introduces multiple acronyms (LDFE, GS2E, LC2E) in close proximity. The full text should ensure each is defined clearly at first use and used consistently throughout.
The phrase 'surpassing the SOTA methods 6.2%, 3.7%, ...' is grammatically incomplete; it should specify whether these are absolute or relative mAP improvements.
Typo: missing space before 'on M3FD' near the end of the abstract ('datasets,respectively' should be 'datasets, respectively').

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity identifiable from abstract: the method is evaluated against external benchmarks and compared to prior SOTA methods, providing independent grounding.

full rationale

This is an abstract-only review, so the full derivation chain, equations, and experimental details are not available for inspection. From the abstract alone, no circularity can be identified. The LDFE block is a proposed architectural contribution (Laplacian Pyramid decomposition, GS2E, LC2E modules) evaluated against six external benchmark datasets (M3FD, DroneVehicle, LLVIP, FLIR-Aligned, KAIST, VEDAI) and compared to prior SOTA methods. The claimed mAP improvements are empirical results measured against external benchmarks, not quantities derived from the paper's own definitions or fitted parameters renamed as predictions. There is no self-definitional structure visible in the abstract, no fitted input being called a prediction, and no self-citation chain invoked as load-bearing evidence. The reader's concern about fair baseline comparisons (training schedules, backbone capacities, input resolutions) is a legitimate correctness risk, but it is not a circularity issue — it concerns experimental confounders, not whether a claimed derivation reduces to its inputs by construction. Without the full text, equations, and tables, no specific circular reduction can be exhibited, and none is apparent from the abstract. The score is therefore 0.

Assumptions & free parameters 5 free parameters · 4 assumptions · 3 invented entities

The axiom ledger is reconstructed from the abstract only. Free parameters and axioms are inferred from the described architecture; the full text would contain additional hyperparameters and design assumptions. The invented entities are the three named modules, none of which have independent evidence beyond the paper's own benchmark results.

free parameters (5)

Laplacian Pyramid levels
The number of decomposition levels used in the Laplacian Pyramid is a design choice that affects feature granularity; value not stated in abstract.
State Space Model dimension
The hidden state dimension of the SSM used in GS2E is a hyperparameter that affects capacity and computational cost; value not stated in abstract.
Cross-modal attention parameters
The attention mechanism that derives suppression signals from the auxiliary modality likely has learnable projection dimensions; not specified in abstract.
LC2E convolution kernel sizes
The 'triple convolution' in LC2E implies specific kernel configurations chosen for local feature extraction; values not stated in abstract.
Training hyperparameters (learning rate, epochs, batch size, augmentation)
Detection mAP is highly sensitive to training recipe; these are free parameters that must be controlled for fair baseline comparison.

assumptions (4)

domain assumption Laplacian Pyramid decomposition cleanly separates features into global and local components suitable for independent processing.
The entire LDFE design rests on the premise that frequency-domain decomposition of CNN features yields semantically meaningful global/local splits. This is plausible but unproven in the abstract.
domain assumption State Space Models effectively capture long-range dependencies in the global feature representations of CNN backbone features.
The GS2E module uses an SSM for global features; the assumption is that SSMs (designed for sequences) transfer effectively to 2D feature map processing.
domain assumption Cross-modal attention from an auxiliary modality can reliably identify and suppress noise in the main modality.
The denoising mechanism assumes that the auxiliary modality contains information that discriminates signal from noise in the main modality. This is a reasonable but unverified assumption.
domain assumption The reported SOTA baselines are fairly comparable (same backbone, training data, schedule).
The mAP improvement claims depend on this axiom. Without the full text, fairness of comparison cannot be confirmed.

invented entities (3)

LDFE block
purpose: Fuses multi-stage dual-stream CNN features via Laplacian decomposition, denoising, and reconstruction.
The LDFE block is the paper's proposed module; its effectiveness is claimed only through the paper's own experiments, which cannot be verified from the abstract.
GS2E (Global State Space Enhancement)
purpose: Denoises and enhances global features using cross-modal attention and a state-space model.
Novel sub-module; no independent evidence available beyond the paper's claims.
LC2E (Local Convolutional Correlation Enhancement)
purpose: Suppresses noise in local features and extracts fine-grained details via triple convolution.
Novel sub-module; no independent evidence available beyond the paper's claims.

how reviews work

0 comments

Cite this review

Pith. "Pith review of LDFE: Laplacian Decoupled Feature Enhancement Block for Dual-Stream CNN-based RGB-IR Object Detection." pith.science (2026). https://pith.science/paper/DZVYWNN6

@misc{pith2026260708076,
  author       = {Pith},
  title        = {Pith review of: LDFE: Laplacian Decoupled Feature Enhancement Block for Dual-Stream CNN-based RGB-IR Object Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DZVYWNN6}},
  note         = {Machine review of arXiv:2607.08076}
}

read the original abstract

The complementary information between RGB and IR images can significantly enhance object detection performance under extreme conditions. Existing methods prefer dual-stream CNN backbones built upon YOLO for feature extraction and focus on the design of feature fusion. In this paper, we introduce the Laplacian Decoupled Feature Enhancement block (LDFE) to fuse features from different stages of the dual-stream CNN backbone. By design, LDFE simultaneously considers the characteristics of modalities and structures for feature fusion by employing global-local decomposition, denoising, fusion, and reconstruction, sequentially. The LDFE first separates features into global and local components based on Laplacian Pyramid, and then performs denoising and fusion based on Global State Space Enhancement module (GS2E) and Local Convolutional Correlation Enhancement module (LC2E) separately. Specifically, the GS2E conducts a two-branch architecture for the main and auxiliary modalities. It dynamically suppresses noise in the main modality through cross-modal attention derived from the auxiliary modality, while employing a State Space Model to capture long-range dependencies within the global feature representations of the main modality. To obtain bidirectional interaction, the two modalities systematically alternate their main/auxiliary roles. Moreover, the LC2E suppresses noise in local features and leverages spatial and channel dimension along with triple convolution to extract fine-grained details for fusion. These innovative designs achieve a significant performance improvement, with mAP surpassing the SOTA methods 6.2%, 3.7%, 4.7%, 2.3%, 4.1% and 2.0% on M3FD, DroneVehicle, LLVIP, FLIR-Aligned, KAIST and VEDAI datasets,respectively.

Figures

Figures reproduced from arXiv: 2607.08076 by the authors.

**Figure 2.** An illustrative diagram of our proposed method, which consists of a dual-stream CNN architecture [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

**Figure 3.** The detailed framework of the Global State Space Enhancement module. M and A denote Main [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗

Figures from the paper (3 more)

**Figure 4.** Figure 4: The detailed structure of the Local Convolutional Correlation Enhancement module with Spatial [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]

**Figure 5.** Figure 5: Heatmap visualization of several cross-modal object detection methods on M [PITH_FULL_IMAGE:figures/full_fig_p019_5.png]

**Figure 6.** Figure 6: Visualization of detection results of several cross-modal object detection methods on M [PITH_FULL_IMAGE:figures/full_fig_p020_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

54 extracted references · 54 canonical work pages

[1]

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

work page 2024
[2]

K. Chen, J. Liu, H. Zhang, Igt: Illumination-guided rgb-t object detection with transformers, Knowledge-Based Systems 268 (2023) 110423

work page 2023
[3]

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

work page 2024
[4]

Zhang, M

J. Zhang, M. Cao, W. Xie, J. Lei, D. Li, W. Huang, Y . Li, X. Yang, E2e-mfd: To- wards end-to-end synchronous multimodal fusion detection, Advances in Neural Information Processing Systems 37 (2024) 52296–52322

work page 2024
[5]

G. Li, Z. Shi, Z. Gu, B. Zheng, H. Zheng, Reference-then-supervision framework for infrared and visible image fusion, Pattern Recognition 158 (2025) 110996

work page 2025
[6]

X. Fei, M. Guo, Y . Li, R. Yu, L. Sun, Acdf-yolo: Attentive and cross-differential fusion network for multimodal remote sensing object detection, Remote Sensing 16 (18) (2024) 3532. 25

work page 2024
[7]

J. Tang, C. Ye, X. Zhou, L. Xu, Yolo-fusion and internet of things: Advancing ob- ject detection in smart transportation, Alexandria Engineering Journal 107 (2024) 1–12

work page 2024
[8]

Y . Chen, B. Wang, W. Zhu, J. Yuan, Rgb-ir yolo combining modality-specific re- construction and information integration, in: 2024 39th Youth Academic Annual Conference of Chinese Association of Automation (Y AC), 2024, pp. 2045–2050

work page 2024

Show all 54 references

[9]

D. Tian, X. Yan, D. Zhou, C. Wang, W. Zhang, Iv-yolo: a lightweight dual-branch object detection network, Sensors 24 (19) (2024) 6181

2024
[10]

T. Zhao, M. Yuan, F. Jiang, N. Wang, X. Wei, Removal and selection: Im- proving rgb-infrared object detection via coarse-to-fine fusion, arXiv preprint arXiv:2401.10731 (2024)

2024
[11]

Y . Wang, J. Liu, J. Wang, L. Yang, B. Dong, Z. Li, Haarfuse: A dual-branch infrared and visible light image fusion network based on haar wavelet transform, Pattern Recognition 164 (2025) 111594

2025
[12]

W. Zhao, Z. Zhao, M. Xu, Y . Ding, J. Gong, Differential multimodal fusion algo- rithm for remote sensing object detection through multi-branch feature extraction, Expert Systems with Applications 265 (2025) 125826

2025
[13]

J. Wang, N. Su, C. Zhao, Y . Yan, S. Feng, Multi-modal object detection method based on dual-branch asymmetric attention backbone and feature fusion pyramid network, Remote Sensing 16 (20) (2024)

2024
[14]

K. Hu, Y . He, Y . Li, J. Zhao, S. Chen, Y . Kang, Ei2det: Edge-guided illumination- aware interactive learning for visible-infrared object detection, IEEE Transactions on Circuits and Systems for Video Technology (2025) 1–1

2025
[15]

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. 26

2024
[16]

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

2025
[17]

M. Zhou, T. Li, C. Qiao, D. Xie, G. Wang, N. Ruan, L. Mei, Y . Yang, Dmm: Disparity-guided multispectral mamba for oriented object detection in remote sensing, arXiv preprint arXiv:2407.08132 (2024)

2024 arXiv
[18]

Cordonnier, A

J.-B. Cordonnier, A. Loukas, M. Jaggi, On the relationship between self-attention and convolutional layers, in: Eighth International Conference on Learning Representations-ICLR 2020, 2020

2020
[19]

X. Ma, Z. Ni, X. Chen, Tinyvim: Frequency decoupling for tiny hybrid vision mamba, arXiv preprint arXiv:2411.17473 (2024)

2024
[20]

Xiong, Z

Z. Xiong, Z. Yao, X. Liu, W. Zhao, J. Cao, X. Wu, Efficient multispectral object detection with attentive feature aggregation leveraging zero-shot implicit illumi- nation guidance, Information Fusion 118 (2025) 102939

2025
[21]

Jocher, A

G. Jocher, A. Chaurasia, J. Qiu, Ultralytics yolov8 (2023). URLhttps://github.com/ultralytics/ultralytics

2023
[22]

L. Tang, X. Xiang, H. Zhang, M. Gong, J. Ma, Divfusion: Darkness-free infrared and visible image fusion, Information Fusion 91 (2023) 477–493

2023
[23]

S. Ding, J. Huang, Z. Cai, Y . Ma, K. Wu, F. Fan, Fiafusion: A feedback-based illumination-adaptive infrared and visible image fusion method, IEEE Sensors Journal (2025)

2025
[24]

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

2024
[25]

X. Sun, Y . Yu, Q. Cheng, Adaptive multimodal feature fusion with frequency domain gate for remote sensing object detection, Remote Sensing Letters 15 (2) (2024) 133–144. 27

2024
[26]

F. Xu, T. Xu, L. Hong, P. Peng, J. Guo, J. Li, Enhanced spectral-spatial fusion network for multispectral object detection in ground-aerial images, IEEE Geo- science and Remote Sensing Letters (2024)

2024
[27]

Z. Zhao, H. Bai, J. Zhang, Y . Zhang, K. Zhang, S. Xu, D. Chen, R. Timofte, L. Van Gool, Equivariant multi-modality image fusion, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2024, pp. 25912–25921

2024
[28]

J. Liu, R. Lin, G. Wu, R. Liu, Z. Luo, X. Fan, Coconet: Coupled contrastive learn- ing network with multi-level feature ensemble for multi-modality image fusion, International Journal of Computer Vision 132 (5) (2024) 1748–1775

2024
[29]

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

2024
[30]

M. Yuan, X. Wei, C 2 former: Calibrated and complementary transformer for rgb- infrared object detection, IEEE Transactions on Geoscience and Remote Sensing (2024)

2024
[31]

H. Zhou, A. Ma, S. Zhou, Y . Niu, Vehicle detection method based on improved yolov8 in rgb-infrared aerial images, in: 2024 China Automation Congress (CAC), 2024, pp. 3003–3008

2024
[32]

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) (2024) 13232–13245

2024
[33]

J. Guo, C. Gao, F. Liu, D. Meng, X. Gao, Damsdet: Dynamic adaptive multispec- tral detection transformer with competitive query selection and adaptive feature fusion, in: European Conference on Computer Vision, 2024, pp. 464–481

2024
[34]

H. Wang, S. Qu, Z. Qiao, X. Liu, Kcdnet: Multimodal object detection in modal information imbalance scenes, IEEE Transactions on Instrumentation and Mea- surement (2024). 28

2024
[35]

X. Luo, G. Fu, J. Yang, Y . Cao, Y . Cao, Multi-modal image fusion via deep laplacian pyramid hybrid network, IEEE Transactions on Circuits and Systems for Video Technology 33 (12) (2023) 7354–7369

2023
[36]

Zhang, H

H. Zhang, H. Shen, Q. Yuan, X. Guan, Multispectral and sar image fusion based on laplacian pyramid and sparse representation, Remote Sensing 14 (4) (2022) 870

2022
[37]

P. J. Burt, E. H. Adelson, The laplacian pyramid as a compact image code, in: Readings in computer vision, Elsevier, 1987, pp. 671–679

1987
[38]

Huang, K

M. Huang, K. Chang, Q. Qin, Y . Tang, G. Li, Conditional laplacian pyramid networks for exposure correction, Signal Processing: Image Communication 134 (2025) 117276

2025
[39]

Li, X.-J

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

2018
[40]

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

2022
[41]

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

2022
[42]

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
[43]

FLIR, Free teledyne flir thermal dataset for algorithm training, Online (April 9 2024)

T. FLIR, Free teledyne flir thermal dataset for algorithm training, Online (April 9 2024)

2024
[44]

Hwang, J

S. Hwang, J. Park, N. Kim, Y . Choi, I. So Kweon, Multispectral pedestrian de- tection: Benchmark dataset and baseline, in: Proceedings of the IEEE conference on computer vision and pattern recognition, 2015, pp. 1037–1045. 29

2015
[45]

Razakarivony, F

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

2016
[46]

Liang, J

M. Liang, J. Hu, C. Bao, H. Feng, F. Deng, T. L. Lam, Explicit attention-enhanced fusion for rgb-thermal perception tasks, IEEE Robotics and Automation Letters 8 (7) (2023) 4060–4067

2023
[47]

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

2025
[48]

C. Chen, J. Qi, X. Liu, K. Bin, R. Fu, X. Hu, P. Zhong, Weakly misalignment-free adaptive feature alignment for uavs-based multimodal object detection, in: Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog- nition, 2024, pp. 26836–26845

2024
[49]

P. Zhao, X. Ye, Z. Du, Object detection in multispectral remote sensing images based on cross-modal cross-attention, Sensors 24 (13) (2024) 4098

2024
[50]

Q. Wang, Y . Chi, T. Shen, J. Song, Z. Zhang, Y . Zhu, Improving rgb-infrared object detection by reducing cross-modality redundancy, Remote Sensing 14 (9) (2022) 2020

2022
[51]

Qingyun, W

F. Qingyun, W. Zhaokui, Cross-modality attentive feature fusion for object de- tection in multispectral remote sensing imagery, Pattern Recognition 130 (2022) 108786

2022
[52]

R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh, D. Batra, Grad- cam: Visual explanations from deep networks via gradient-based localization, in: Proceedings of the IEEE international conference on computer vision, 2017, pp. 618–626. 30

2017
[53]

S. A. Deevi, C. Lee, L. Gan, S. Nagesh, G. Pandey, S.-J. Chung, Rgb-x object detection via scene-specific fusion modules, in: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2024, pp. 7366–7375

2024
[54]

Zhang, H

J. Zhang, H. Liu, K. Yang, X. Hu, R. Liu, R. Stiefelhagen, Cmx: Cross-modal fusion for rgb-x semantic segmentation with transformers, IEEE Transactions on intelligent transportation systems 24 (12) (2023) 14679–14694. 31

2023

Pith tools

Reviewed July 10, 2026 · model on record in the stance chip above.

[1] [1]

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

work page 2024

[2] [2]

K. Chen, J. Liu, H. Zhang, Igt: Illumination-guided rgb-t object detection with transformers, Knowledge-Based Systems 268 (2023) 110423

work page 2023

[3] [3]

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

work page 2024

[4] [4]

Zhang, M

J. Zhang, M. Cao, W. Xie, J. Lei, D. Li, W. Huang, Y . Li, X. Yang, E2e-mfd: To- wards end-to-end synchronous multimodal fusion detection, Advances in Neural Information Processing Systems 37 (2024) 52296–52322

work page 2024

[5] [5]

G. Li, Z. Shi, Z. Gu, B. Zheng, H. Zheng, Reference-then-supervision framework for infrared and visible image fusion, Pattern Recognition 158 (2025) 110996

work page 2025

[6] [6]

X. Fei, M. Guo, Y . Li, R. Yu, L. Sun, Acdf-yolo: Attentive and cross-differential fusion network for multimodal remote sensing object detection, Remote Sensing 16 (18) (2024) 3532. 25

work page 2024

[7] [7]

J. Tang, C. Ye, X. Zhou, L. Xu, Yolo-fusion and internet of things: Advancing ob- ject detection in smart transportation, Alexandria Engineering Journal 107 (2024) 1–12

work page 2024

[8] [8]

Y . Chen, B. Wang, W. Zhu, J. Yuan, Rgb-ir yolo combining modality-specific re- construction and information integration, in: 2024 39th Youth Academic Annual Conference of Chinese Association of Automation (Y AC), 2024, pp. 2045–2050

work page 2024