Pith. sign in

REVIEW 4 major objections 7 minor 49 references

Make Both Ends Meet: A Synergistic Optimization Infrared Small Target Detection with Streamlined Computational Overhead

T0 review · 4 major / 7 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read The paper argues that a lightweight YOLOv8-n-based detector with four efficiency modules matches or beats segmentation-based methods on infrared small target detection at a fraction of the compute.

desk verdict A competent lightweight IRSTD detector with public code, but the SOTA claim rests on an undisclosed validation protocol that the paper needs to clarify. read the letter →

arxiv 2504.21581 v3 pith:73JK4IAC submitted 2025-04-30 eess.IV

Yuxin Jing , Yuchen Zheng , Jufeng Zhao , Guangmang Cui , Tianpei Zhang This is my paper · ORCID
classification eess.IV
keywords infraredsmalltargetdetectionlightweightnetworkYOLOv8MobileInvertedBottleneckConvolutionPartialVariableKernelGSConvreal-time
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 tries to show that infrared small target detection (IRSTD) can be reformulated as bounding-box detection rather than pixel-level segmentation without sacrificing accuracy. It introduces LE-IRSTD, built on the YOLOv8-n detector, whose backbone and neck replace heavy standard convolutions with Mobile Inverted Bottleneck blocks, partial-convolution bottleneck blocks, an attention-based variable-kernel convolution stem, and channel-shuffle convolution. On three public IRSTD benchmarks the model reports mAP@50 of 95.4, 95.3, and 94.1 with 2.6M parameters and 7.4 GFLOPs, surpassing several larger segmentation and detection models. The significance is practical: if the claim holds, real-time infrared surveillance and targeting systems can get high detection accuracy on small, dim targets without the compute budget segmentation methods demand.

What carries the argument

The load-bearing machinery is the LE-IRSTD architecture itself, a YOLOv8-n derivative with four substitutions. MBConvblock expands channels $6\times$ with a $1\times 1$ convolution, applies a $3\times 3$ depthwise convolution, adds CBAM attention and dropout, and uses a residual connection when channels match; it replaces repeated C2f bottlenecks to cut depth. BSblock uses Partial Convolution (PConv) on one quarter of the channels, concatenates the untouched channels, passes through an MLP with dropout, and residual-adds; the paper says this cuts memory access to 25% and FLOPs to $1/16$ of standard convolution at ratio $r = 1/4$. AVCStem fuses parallel spatial and channel branches and uses VKConv to learn per-pixel offsets (scaled by $\alpha = 0.1$) that deform the $5\times 5$ sampling grid to match target shape. GSConv runs a $3\times 3$ convolution, then a $3\times 3$ depthwise convolution, concatenates, and channel-shuffles the result. The loss is a weighted sum $L_{\mathrm{all}} = 0.02 L_{\mathrm{BCE}} + 0.49 L_{\mathrm{CIoU}} + 0.49 L_{\mathrm{DFL}}$, deliberately down-weighting classification because the task has one class.

What would settle it

Run a validation-only hyperparameter search over the loss weights and MBConv width settings, lock the chosen values, and then evaluate on the test split exactly once; if the resulting IRSTD-1K mAP falls below the reported 95.4 or below OSCAR's 95.0, the claimed state-of-the-art result would not survive an unbiased protocol.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is that the accuracy/efficiency trade-off in IRSTD is not fixed: a careful choice of lightweight blocks can improve both at once. LE-IRSTD combines four modules — MBConvblock for expand-then-depthwise feature extraction, BSblock which applies partial convolution to a quarter of the channels and preserves the rest, AVCStem which adapts convolution kernel shapes via variable kernel convolution (VKConv) to fit irregular small targets, and GSConv which interleaves standard and depthwise features with channel shuffle. With these, the model reports 89.8 precision, 91.2 recall, 90.5 F1, and 95.4 mAP@50 (mean average precision at IoU threshold 0.5) on IRSTD-1K, 95.3 mAP on NUAA-SIRST, and 94.1 mAP on NUDT-SIRST, while reducing parameters from 3.0M to 2.6M and FLOPs from 8.2G to 7.4G relative to the YOLOv8-n baseline. The paper further claims that the bounding-box formulation avoids the IoU 'cliff drop' that pixel-level annotation ambiguity causes for small targets.

Load-bearing premise

The load-bearing premise is that the model's settings, especially the loss weights $\lambda_1 = 0.02$, $\lambda_2 = 0.49$, $\lambda_3 = 0.49$, were chosen on a validation split and not by looking at test-set labels; the paper does not state this, and the ablation tables report the same final mAP as the main comparison.

Editorial extensions

If this is right

  • If LE-IRSTD's numbers are reproducible, bounding-box detection becomes a viable default paradigm for IRSTD, shifting the field's focus from segmentation refinement to detector efficiency.
  • The 7.4 GFLOPs and 2.6M-parameter budget makes real-time onboard deployment plausible on edge GPUs or embedded devices, which the much heavier segmentation baselines cannot offer.
  • The loss-weighting result implies that for single-class small-target detection, allocating nearly all loss weight to box regression and localization (CIoU and DFL) rather than classification improves mAP.
  • The reported mNoCoAP improvements (e.g., 87.2 versus 85.0 for OSCAR on IRSTD-1K) indicate that the gain is not only box overlap but also target contrast and centroid localization, a metric designed to be paradigm-independent.
  • The four modules are generic enough that swapping C2f bottlenecks for MBConv/BS blocks and standard neck convolutions for GSConv could reduce FLOPs in other YOLO-family small-object detectors.

Reading between the lines

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

  • A natural extension is to test the same four-module recipe on non-infrared small-object benchmarks such as aerial, drone, or maritime imagery; the paper's mechanism is generic, so if the gains persist the method is a general small-target efficiency recipe rather than an infrared-specific one.
  • The paper's use of mNoCoAP suggests a testable prediction: bounding-box detectors trained with heavy regression loss will show smaller mNoCoAP-to-mAP gaps than segmentation-based methods, because they optimize centroid localization directly.
  • The hyperparameter search over loss weights (Table VII) could be rerun under strict validation-only selection; confirming the reported 95.4 mAP under that protocol would settle whether the state-of-the-art claim depends on test-set feedback.
  • The PConv ratio $r = 1/4$ and MBConv expansion factor $6$ are reported as chosen values, not proven optima; a validation-set sensitivity sweep over these two numbers would reveal how sharply performance depends on them.
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 / 7 minor

Summary. The paper proposes LE-IRSTD, a lightweight infrared small target detection framework built on YOLOv8-n, where the backbone C2f modules are replaced by Mobile Inverted Bottleneck (MBConv) and Bottleneck Structure (BS) blocks, an Attention-based Variable Convolution Stem (AVCStem) with variable kernel convolution is introduced in the neck, and GSConv is used for downsampling. The loss is a weighted combination of BCE, CIoU, and a component the paper calls DFL. Experiments on IRSTD-1K, NUAA-SIRST, and NUDT-SIRST report 2.6M parameters, 7.4 GFLOPs, and mAP@50 of 95.4, 95.3, and 94.1, respectively, outperforming segmentation- and detection-based baselines including OSCAR, and the authors release the code. The paper argues that bounding-box regression is better suited to IRSTD than semantic segmentation because pixel-level IoU is overly sensitive to annotation ambiguity for tiny targets.

Significance. If the reported results hold, the paper would demonstrate that a compact YOLOv8 variant can beat much heavier segmentation-based IRSTD models with roughly 16x fewer parameters and 9x fewer FLOPs, which is a practically valuable result. The conceptual argument for bounding-box modeling of IRSTD is well motivated, and the lightweight design based on existing blocks is plausible. Concrete strengths include the public code release, the use of the mNoCoAP metric to compare across detection and segmentation paradigms, and systematic module-wise ablations. The central caveat is that the headline numbers currently rest on an undocumented hyperparameter-selection protocol and on single-run comparisons, so the quantitative claims are plausible but not yet established.

major comments (4)
  1. [IV.A.1, IV.C.1, Tables III and VII] Section IV.A.1 defines a 60/20/20 train/validation/test split, but the paper never states that the architecture and loss-weight ablations in Tables III and VII were evaluated on the validation split rather than the test split. The final rows of both ablation tables report exactly the IRSTD-1K mAP of 95.4 that also appears in the final SOTA row of Table I, which is consistent with configuration selection being driven by test-set feedback. Because the reported margin over the runner-up (OSCAR, 95.0 in Table I) is only 0.4 mAP, tuning on test labels could plausibly account for the claimed advantage. The authors must state which split was used for all ablations and, if test labels were consulted during development, re-run the selection on validation data; otherwise the headline SOTA claim is not supported. In addition, all reported numbers appear to come from a single run, so the authors should provide multiple-seed means with standard deviations or confidence intervals for the central comparisons.
  2. [IV.B, IV.A.3] Section IV.B states that all comparison methods were retrained for 300 epochs with all other parameters at their default values, whereas LE-IRSTD's own configuration was selected through the ablations in Tables III-VII (loss weights 0.02/0.49/0.49, 6x expansion, 3x3 kernel, depths 1 and 2). Comparing a tuned model against default-configured baselines biases the comparison in favor of LE-IRSTD, so the claimed superiority over OSCAR and EFLNet is not an apples-to-apples result as presented. The authors should either give the baselines a comparable tuning budget, such as a small search over loss weights or resolution, or explicitly frame the comparison as tuned-model-versus-default-baselines and temper the SOTA claim accordingly.
  3. [III.C, Eq. (17)] Equation (17), labeled 'Dynamic Focal Loss (DFL),' reads L_DFL = -alpha(1-p)^gamma log(p), which is the standard Focal Loss formula; the Distribution Focal Loss used in YOLOv8 has a different form tied to soft labels over the box's corner distribution, and the accompanying text describing alpha as a class balance factor and gamma as a focusing factor confirms that the focal-loss formula, not DFL, is written down. If the implementation actually uses YOLOv8's DFL, the equation does not describe the implemented loss, and if focal loss is used, the method deviates from YOLOv8 and the loss-weight ablation should be interpreted accordingly. The notation also collides, since alpha denotes the VKConv scaling factor in Eq. (8), the CIoU weighting term in Eq. (15), and the focal balance factor in Eq. (17); please correct the formula, rename the symbols, and state which loss is actually implemented.
  4. [Tables II, V, VII; Section IV.C.6] The ablation tables supporting the final configuration contain internal inconsistencies that need correction or explanation. In Table V, the SC row reports P=83.37, R=85.75, F1=82.77, but the harmonic mean of 83.37 and 85.75 is approximately 84.5, so the reported F1 is arithmetically impossible, and the VKConv row's F1 of 90.8 also differs from the harmonic mean of its P and R values (90.5). In Table II, row (c) reports F1=85.2 for P=88.2, R=81.7, whose harmonic mean is 84.8. In Table VII, the row (0.08, 0.46, 0.46) reports mAP 82.3, a large non-monotonic drop from the neighboring rows (90.7 and 87.5 mAP), which looks like a typo or an unstable run, and Section IV.C.6 states the optimum is lambda1=0.02, lambda2=0.02, lambda3=0.49, contradicting the table's lambda2=lambda3=0.49. These inconsistencies matter because Tables III and VII are the evidence that the reported configuration was chosen for accuracy rather than by chance.
minor comments (7)
  1. [IV.A.3] The training description states 'an initial learning rate of 0.001 and a final learning rate of 0.5' with cosine annealing, which would increase the learning rate over training and is presumably a typo; it also lists 'momentum of 0.937', an SGD hyperparameter, while the stated optimizer is Adam. Please correct the hyperparameter description to match the implementation.
  2. [III.A.3, III.B.4, Table II caption, Fig. 3] The module is called 'Global Shuffle Convolution' in Sections III.A.3 and III.B.4 but 'Ghost Shuffle Convolution' in the Index Terms and in Contribution 3, and 'Attention-based Variable Convolution Stem' appears as 'Adaptive Vision Convolution Stem' in the Table II caption; Figure 3's caption also references an undefined 'Faststem' backbone and 'AVSFPN' neck. Please unify all module and component names.
  3. [Eq. (13)] Equation (13) for the BCE loss has unbalanced parentheses as printed; the intended formula is L_BCE = -(y log(p(x)) + (1-y) log(1-p(x))).
  4. [Introduction, II.B.2, Reference [17]] The same method is cited as 'YOLO-SDLUWD' in the Introduction and as 'YOLO-SLWD' in Section II.B.2; please use a single consistent name.
  5. [Abstract, I, IV.B] The abstract and introduction claim real-time suitability, but no inference latency or FPS is reported anywhere in the experiments; either add latency measurements on the stated hardware or qualify the real-time claim.
  6. [IV.A.1] Section IV.A.1 describes NUDT-SIRST as containing '1327 training and validation examples', which is ambiguous about the total image count used for the 60/20/20 split; please state the total size of each dataset and the exact split sizes.
  7. [Abstract, Index Terms] The text would benefit from a language pass: the abstract contains typos such as 'feature stackin' and 'lightwight', and sentences like 'while most existing methods model IRSTD as a semantic segmentation task, but they suffer from two critical drawbacks' are ungrammatical.

Circularity Check

0 steps flagged · score 0.0 of 10

No construction-level circularity: LE-IRSTD's mAP claims rest on external benchmarks and standard loss design; the hyperparameter-selection ambiguity is a soundness/protocol concern, not a self-referential derivation.

full rationale

The paper's claimed derivation chain is an empirical architecture-engineering report, not a formal derivation of its target metric. Each building block (MBConvblock, BSblock/PConv, AVCStem/VKConv, GSConv) is imported from cited external components (e.g., CBAM [32], LDConv [34], partial convolution [33]) and combined with YOLOv8n; no module is defined in terms of the reported mAP. Eq. (18) is the standard YOLO loss (BCE + CIoU + DFL), and the reported mAP is obtained by training and evaluating on public splits. Section IV.A.1 states: "Each dataset was divided into a training set (60%), validation set (20%), and test set (20%)", so the metric is not an algebraic rewrite of the loss or of any fitted parameter. The reference list contains no self-citations by the authors, so the self-citation and imported-uniqueness patterns do not apply. The strongest caveat is procedural: the ablation tables (III, V, VI, VII) end at the same IRSTD-1K mAP (95.4) as the final comparison row in Table I, and Section IV.C does not explicitly say that the searched widths, depths, kernel sizes, and loss weights were selected on the validation split rather than the test split. That is a test-set-selection / overfitting concern about whether 95.4 is an unbiased estimate, not a circularity in the sense of a quantity being equivalent to its own inputs by definition. It should be evaluated under soundness and reproducibility, not as a circular derivation. No equation reduces to another by construction, and no fitted parameter is renamed as a prediction. Therefore the circularity score is 0.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical or theoretical entities. It recombines known architectural blocks and tunes several hyperparameters on the evaluation benchmark, which is the main source of fitted values and the key methodological concern.

free parameters (7)
  • loss weights lambda1, lambda2, lambda3 = 0.02, 0.49, 0.49
    Selected by ablation in Table VII on IRSTD-1K; the paper does not state whether this was the validation or test split.
  • MBConv expansion ratio = 6
    Chosen via width ablation in Table III on IRSTD-1K.
  • MBConv kernel size = 3x3
    Chosen via kernel ablation in Table III on IRSTD-1K.
  • MBConv depth = 1,2
    Chosen via depth ablation in Table III on IRSTD-1K.
  • VKConv scaling factor alpha = 0.1
    Stated default in Eq. 8; no ablation is reported.
  • dropout probability = p=0.2
    Stated in BSblock description; no ablation is reported.
  • PConv compression ratio r = 1/4
    Taken from the FasterNet prior work; no ablation in this paper.
assumptions (4)
  • domain assumption Bounding box regression is a more suitable formulation than semantic segmentation for IRSTD.
    The paper's motivation in the Introduction and Related Work; the strongest claim depends on this modeling choice.
  • domain assumption The three public datasets (IRSTD-1K, NUAA-SIRST, NUDT-SIRST) with YOLO-format bounding box annotations provide reliable ground truth.
    Section IV.A uses these datasets to evaluate all methods; if annotations are noisy, comparisons are affected.
  • domain assumption Default hyperparameters of the baseline methods were sufficient for a fair comparison.
    Section IV.B states all models were retrained for 300 epochs with default parameters; no per-method tuning is described.
  • domain assumption mAP@0.5 and mNoCoAP are appropriate metrics for assessing IRSTD performance.
    Section IV.A.2 defines these metrics; the claim of superiority relies on them.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Make Both Ends Meet: A Synergistic Optimization Infrared Small Target Detection with Streamlined Computational Overhead." pith.science (2026). https://pith.science/paper/73JK4IAC

@misc{pith2026250421581,
  author       = {Pith},
  title        = {Pith review of: Make Both Ends Meet: A Synergistic Optimization Infrared Small Target Detection with Streamlined Computational Overhead},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/73JK4IAC}},
  note         = {Machine review of arXiv:2504.21581}
}
read the original abstract

Infrared small target detection(IRSTD) is widely recognized as a challenging task due to the inherent limitations of infrared imaging, including low signal-to-noise ratios, lack of texture details, and complex background interference. While most existing methods model IRSTD as a semantic segmentation task, but they suffer from two critical drawbacks: (1)blurred target boundaries caused by long-distance imaging dispersion; and (2) excessive computational overhead due to indiscriminate feature stackin. To address these issues, we propose the Lightweight Efficiency Infrared Small Target Detection (LE-IRSTD), a lightweight and efficient framework based on YOLOv8n, with following key innovations. Firstly, we identify that the multiple bottleneck structures within the C2f component of the YOLOv8-n backbone contribute to an increased computational burden. Therefore, we implement the Mobile Inverted Bottleneck Convolution block (MBConvblock) and Bottleneck Structure block (BSblock) in the backbone, effectively balancing the trade-off between computational efficiency and the extraction of deep semantic information. Secondly, we introduce the Attention-based Variable Convolution Stem (AVCStem) structure, substituting the final convolution with Variable Kernel Convolution (VKConv), which allows for adaptive convolutional kernels that can transform into various shapes, facilitating the receptive field for the extraction of targets. Finally, we employ Global Shuffle Convolution (GSConv) to shuffle the channel dimension features obtained from different convolutional approaches, thereby enhancing the robustness and generalization capabilities of our method. Experimental results demonstrate that our LE-IRSTD method achieves compelling results in both accuracy and lightweight performance, outperforming several state-of-the-art deep learning methods.

Figures

Figures reproduced from arXiv: 2504.21581 by the authors.

Figure 1
Figure 1. The different small targets exist in the infrared image, The [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Disadvantages of IoU for small targets. (a) Sensitivity analysis of IoU [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. The overall framework of LE-IRSTD is shown in the figure, which includes the feature extraction backbone Faststem, the improved neck AVSFPN, [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (7 more)
Figure 5
Figure 5. Figure 5: Overview of Bottleneck Structure block. PConv stands for Partial [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 4
Figure 4. Figure 4: Structure of Mobile Inverted Bottleneck Convolution. Box 1 represents [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 7
Figure 7. Figure 7: The orginal coordinate generation algorithm is illustrated. The original [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]
Figure 8
Figure 8. Figure 8: The structure of GSConv: the input first undergoes "Convolution, [PITH_FULL_IMAGE:figures/full_fig_p006_8.png]
Figure 9
Figure 9. Figure 9: Different annotation forms for the current infrared small target public [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]
Figure 10
Figure 10. Figure 10: Partial visual results obtained from different methods on the IRSTD-1k dataset are shown. In the figure, the red, orange, and purple circles represent [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]
Figure 11
Figure 11. Figure 11: The improved gradmap effects of adding different model components are displayed. The more prominent the gradmap, the stronger the ability to [PITH_FULL_IMAGE:figures/full_fig_p011_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 44 canonical work pages

  1. [1]

    Mtu-net: Multilevel transunet for space-based infrared tiny ship detection,

    T. Wu, B. Li, Y . Luo, Y . Wang, C. Xiao, T. Liu, J. Yang, W. An, and Y . Guo, “Mtu-net: Multilevel transunet for space-based infrared tiny ship detection,” IEEE Transactions on Geoscience and Remote Sensing , vol. 61, pp. 1–15, 2023. 1

    work page 2023

  2. [2]

    Smpisd-mtpnet: Scene semantic prior- assisted infrared ship detection using multi-task perception networks,

    C. Hu, X. Dong, Y . Huang, L. Wang, L. Xu, T. Pu, and Z. Peng, “Smpisd-mtpnet: Scene semantic prior- assisted infrared ship detection using multi-task perception networks,” IEEE Transactions on Geoscience and Remote Sensing, 2024. 1

    2024

  3. [3]

    Small infrared target detection based on weighted local differ- ence measure,

    H. Deng, X. Sun, M. Liu, C. Ye, and X. Zhou, “Small infrared target detection based on weighted local differ- ence measure,” IEEE Transactions on Geoscience and Remote Sensing , vol. 54, no. 7, pp. 4204–4214, 2016. 1

    work page 2016

  4. [4]

    Infrared sensors for ballistic missile defense,

    M. Tidrow and W. Dyer, “Infrared sensors for ballistic missile defense,” Infrared Physics & Technology , vol. 42, no. 3-5, pp. 333–336, 2001. 1

    work page 2001

  5. [5]

    Infrared thermography for condition monitoring–a review,

    S. Bagavathiappan, B. B. Lahiri, T. Saravanan, J. Philip, and T. Jayakumar, “Infrared thermography for condition monitoring–a review,” Infrared Physics & Technology , vol. 60, pp. 35–55, 2013. 1

    work page 2013

  6. [6]

    Infrared small target segmentation networks: A survey,

    R. Kou, C. Wang, Z. Peng, Z. Zhao, Y . Chen, J. Han, F. Huang, Y . Yu, and Q. Fu, “Infrared small target segmentation networks: A survey,” Pattern Recognition, vol. 143, p. 109788, 2023. 1

    2023

  7. [7]

    Towards accurate infrared small target detection via edge- aware gated transformer,

    Y . Zhu, Y . Ma, F. Fan, J. Huang, K. Wu, and G. Wang, “Towards accurate infrared small target detection via edge- aware gated transformer,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , 2024. 3

    work page 2024

  8. [8]

    ISNET: Shape matters for infrared small target detection,

    M. Zhang, R. Zhang, Y . Yang, H. Bai, J. Zhang, and J. Guo, “ISNET: Shape matters for infrared small target detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022, pp. 877–886. 1, 7, 8, 9

    work page 2022

Show all 49 references
  1. [9]

    Gradient is all you need: Gradient-based attention fusion for infrared small target detection,

    C. Hu, Y . Huang, K. Li, L. Zhang, Y . Zhu, Y . Peng, T. Pu, and Z. Peng, “Gradient is all you need: Gradient-based attention fusion for infrared small target detection,” 2024. [Online]. Available: https://arxiv.org/abs/2409.19599 1

    2024

  2. [10]

    Feature gradients: Scalable feature selection via discrete relaxation,

    R. Sheth and N. Fusi, “Feature gradients: Scalable feature selection via discrete relaxation,” arXiv preprint arXiv:1908.10382, 2019

    1908 arXiv

  3. [11]

    M4net: Multi-level multi-patch multi-receptive multi-dimensional attention network for infrared small target detection,

    F. Zhang, H. Hu, B. Zou, and M. Luo, “M4net: Multi-level multi-patch multi-receptive multi-dimensional attention network for infrared small target detection,” Neural Networks, vol. 183, p. 107026, 2025. 1

    2025

  4. [12]

    Attentional local contrast networks for infrared small target detection,

    Y . Dai, Y . Wu,F. Zhou, and K. Barnard, “Attentional local contrast networks for infrared small target detection,” IEEE transactions on geoscience and remote sensing , vol. 59, no. 11, pp. 9813–9824, 2021. 1, 8, 9

    2021

  5. [13]

    A local contrast method for small infrared target detection,

    C. P. Chen, H. Li, Y . Wei, T. Xia, and Y . Y . Tang, “A local contrast method for small infrared target detection,” IEEE transactions on geoscience and remote sensing , vol. 52, no. 1, pp. 574–581, 2013. 3

    2013

  6. [14]

    Hierarchical attention-guided multiscale aggregation net- 13 work for infrared small target detection,

    S. Zhong, H. Zhou, Z. Zheng, Z. Ma, F. Zhang et al. , “Hierarchical attention-guided multiscale aggregation net- 13 work for infrared small target detection,” Neural Networks, vol. 171, pp. 485–496, 2024. 2

    2024

  7. [15]

    Single- frame infrared small-target detection: A survey,

    M. Zhao, W. Li, L. Li, J. Hu, P. Ma, and R. Tao, “Single- frame infrared small-target detection: A survey,” IEEE Geoscience and Remote Sensing Magazine , vol. 10, no. 2, pp. 87–119, 2022. 2

    2022

  8. [16]

    One-stage cascade refinement networks for infrared small target detection,

    Y . Dai, X. Li, F. Zhou, Y . Qian, Y . Chen, and J. Yang, “One-stage cascade refinement networks for infrared small target detection,” IEEE Transactions on Geoscience and Remote Sensing , vol. 61, pp. 1–17, 2023. 2, 3, 8, 9

    2023

  9. [17]

    Yolo-sdluwd: Yolov7- based small target detection network for infrared images in complex backgrounds,

    J. Zhu, C. Qin, and D. Choi, “Yolo-sdluwd: Yolov7- based small target detection network for infrared images in complex backgrounds,” Digital Communications and Networks, 2023. 2, 3

    2023

  10. [18]

    Attention is all you need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” Advances in neural information processing systems, vol. 30, 2017. 2

    2017

  11. [19]

    Eflnet: Enhancing feature learning network for infrared small target detection,

    B. Yang, X. Zhang, J. Zhang, J. Luo, M. Zhou, and Y . Pi, “Eflnet: Enhancing feature learning network for infrared small target detection,” IEEE Transactions on Geoscience and Remote Sensing , vol. 62, pp. 1–11, 2024. 2, 3, 8, 9

    2024

  12. [20]

    Max-mean and max-median filters for detection of small targets,

    S. D. Deshpande, M. H. Er, R. Venkateswarlu, and P. Chan, “Max-mean and max-median filters for detection of small targets,” in Signal and Data Processing of Small Targets 1999, vol. 3809. SPIE, 1999, pp. 74–83. 3

    1999

  13. [21]

    Analysis of new top-hat transfor- mation and the application for infrared dim small target detection,

    X. Bai and F. Zhou, “Analysis of new top-hat transfor- mation and the application for infrared dim small target detection,” Pattern Recognition, vol. 43, no. 6, pp. 2145– 2156, 2010. 3, 8, 9

    2010

  14. [22]

    Infrared small and dim target detection with transformer under complex backgrounds,

    F. Liu, C. Gao, F. Chen, D. Meng, W. Zuo, and X. Gao, “Infrared small and dim target detection with transformer under complex backgrounds,” IEEE Transactions on Image Processing, vol. 32, pp. 5921–5932, 2023. 3

    2023

  15. [23]

    Infrared small target detection based on the weighted strengthened local contrast measure,

    J. Han, S. Moradi, I. Faramarzi, H. Zhang, Q. Zhao, X. Zhang, and N. Li, “Infrared small target detection based on the weighted strengthened local contrast measure,” IEEE Geoscience and Remote Sensing Letters , vol. 18, no. 9, pp. 1670–1674, 2020. 3

    2020

  16. [24]

    A local contrast method for infrared small-target detection utilizing a tri-layer window,

    J. Han, S. Moradi, I. Faramarzi, C. Liu, H. Zhang, and Q. Zhao, “A local contrast method for infrared small-target detection utilizing a tri-layer window,” IEEE Geoscience and Remote Sensing Letters , vol. 17, no. 10, pp. 1822– 1826, 2019. 3, 8, 9

    2019

  17. [25]

    Infrared small target detection by density peaks searching and maximum-gray region growing,

    S. Huang, Z. Peng, Z. Wang, X. Wang, and M. Li, “Infrared small target detection by density peaks searching and maximum-gray region growing,” IEEE Geoscience and Remote Sensing Letters , vol. 16, no. 12, pp. 1919– 1923, 2019. 3

    1919

  18. [26]

    Reweighted infrared patch-tensor model with both nonlocal and local priors for single- frame small target detection,

    Y . Dai and Y . Wu, “Reweighted infrared patch-tensor model with both nonlocal and local priors for single- frame small target detection,” IEEE journal of selected topics in applied earth observations and remote sensing , vol. 10, no. 8, pp. 3752–3767, 2017. 3, 8, 9

    2017

  19. [27]

    Towards robust infrared small target detection via frequency and spatial feature fusion,

    Y . Zhu, Y . Ma, F. Fan, J. Huang, Y . Yao, X. Zhou, and R. Huang, “Towards robust infrared small target detection via frequency and spatial feature fusion,” IEEE Transactions on Geoscience and Remote Sensing , 2025. 3

    2025

  20. [28]

    Dnanet: Dense nested attention network for single image dehazing,

    D. Ren, J. Li, M. Han, and M. Shu, “Dnanet: Dense nested attention network for single image dehazing,” in ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2021, pp. 2035–2039. 3

    2021

  21. [29]

    Acm-net: Action context modeling network for weakly- supervised temporal action localization,

    S. Qu, G. Chen, Z. Li, L. Zhang, F. Lu, and A. Knoll, “Acm-net: Action context modeling network for weakly- supervised temporal action localization,” arXiv preprint arXiv:2104.02967, 2021. 3

    2021 arXiv

  22. [30]

    Background semantics matter: Cross-task feature exchange network for clustered infrared small target detection with sky-annotated dataset,

    M. Xiao, Q. Dai, Y . Zhu, K. Guo, H. Wang, X. Shu, J. Yang, and Y . Dai, “Background semantics matter: Cross-task feature exchange network for clustered infrared small target detection with sky-annotated dataset,” arXiv preprint arXiv:2407.20078, 2024. 3

    2024

  23. [31]

    Focal loss for dense object detection,

    T.-Y . Ross and G. Dollár, “Focal loss for dense object detection,” in proceedings of the IEEE conference on computer vision and pattern recognition , 2017, pp. 2980–

    2017

  24. [32]

    Cbam: Convolutional block attention module,

    S. Woo, J. Park, J.-Y . Lee, and I. S. Kweon, “Cbam: Convolutional block attention module,” in Proceedings of the European conference on computer vision (ECCV) , 2018, pp. 3–19. 4

    2018

  25. [33]

    Run, don’t walk: chasing higher flops for faster neural networks,

    J. Chen, S.-h. Kao, H. He, W. Zhuo, S. Wen, C.-H. Lee, and S.-H. G. Chan, “Run, don’t walk: chasing higher flops for faster neural networks,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023, pp. 12 021–12 031. 5

    2023

  26. [34]

    Ldconv: Linear deformable convolution for improving convolutional neural networks,

    X. Zhang, Y . Song, T. Song, D. Yang, Y . Ye, J. Zhou, and L. Zhang, “Ldconv: Linear deformable convolution for improving convolutional neural networks,” Image and Vision Computing, vol. 149, p. 105190, 2024. 6

    2024

  27. [35]

    Dense nested attention network for infrared small target detection,

    B. Li, C. Xiao, L. Wang, Y . Wang, Z. Lin, M. Li, W. An, and Y . Guo, “Dense nested attention network for infrared small target detection,” IEEE Transactions on Image Processing, vol. 32, pp. 1745–1758, 2022. 6, 7, 8, 9

    2022

  28. [36]

    Asymmetric contextual modulation for infrared small target detection,

    Y . Dai, Y . Wu, F. Zhou, and K. Barnard, “Asymmetric contextual modulation for infrared small target detection,” in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision , 2021, pp. 950–959. 7, 8, 9

    2021

  29. [37]

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

    R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh, and 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. 7, 9

    2017

  30. [38]

    Adam: A method for stochastic optimization,

    D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” arXiv preprint arXiv:1412.6980 , 2014. 8

    2014 arXiv

  31. [39]

    Infrared small target detection based on facet kernel and random walker,

    Y . Qin, L. Bruzzone, C. Gao, and B. Li, “Infrared small target detection based on facet kernel and random walker,” IEEE Transactions on Geoscience and Remote Sensing , vol. 57, no. 9, pp. 7104–7118, 2019. 8, 9

    2019

  32. [40]

    Multiscale patch-based contrast measure for small infrared target detection,

    Y . Wei, X. You, and H. Li, “Multiscale patch-based contrast measure for small infrared target detection,” Pattern Recognition, vol. 58, pp. 216–226, 2016. 8, 9

    2016

  33. [41]

    Infrared patch-image model for small target detection in a single image,

    C. Gao, D. Meng, Y . Yang, Y . Wang, X. Zhou, and A. G. Hauptmann, “Infrared patch-image model for small target detection in a single image,” IEEE transactions on image 14 processing, vol. 22, no. 12, pp. 4996–5009, 2013. 8, 9

    2013

  34. [42]

    Agpcnet: Attention- guided pyramid context networks for infrared small target detection,

    T. Zhang, S. Cao, T. Pu, and Z. Peng, “Agpcnet: Attention- guided pyramid context networks for infrared small target detection,” arXiv preprint arXiv:2111.03580 , 2021. 8, 9

    2021 arXiv

  35. [43]

    Sctransnet: Spatial-channel cross transformer network for infrared small target detection,

    S. Yuan, H. Qin, X. Yan, N. Akhtar, and A. Mian, “Sctransnet: Spatial-channel cross transformer network for infrared small target detection,” IEEE Transactions on Geoscience and Remote Sensing , 2024. 8, 9

    2024

  36. [44]

    Rich feature hierarchies for accurate object detection and semantic segmentation,

    R. Girshick, J. Donahue, T. Darrell, and J. Malik, “Rich feature hierarchies for accurate object detection and semantic segmentation,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2014, pp. 580–587. 8, 9

    2014

  37. [45]

    Faster r-cnn: Towards real-time object detection with region proposal networks,

    S. Ren, K. He, R. Girshick, and J. Sun, “Faster r-cnn: Towards real-time object detection with region proposal networks,” IEEE transactions on pattern analysis and machine intelligence, vol. 39, no. 6, pp. 1137–1149, 2016. 8, 9

    2016

  38. [46]

    Amflw-yolo: A lightweight network for remote sensing image detection based on attention mechanism and multi-scale feature fusion,

    G. Peng, Z. Yang, S. Wang, and Y . Zhou, “Amflw-yolo: A lightweight network for remote sensing image detection based on attention mechanism and multi-scale feature fusion,” IEEE Transactions on Geoscience and Remote Sensing, 2023. 8, 9

    2023

  39. [47]

    Det-yolo: An innovative high- performance model for detecting military aircraft in remote sensing images,

    X. Chen, H. Jiang, H. Zheng, J. Yang, R. Liang, D. Xiang, H. Cheng, and Z. Jiang, “Det-yolo: An innovative high- performance model for detecting military aircraft in remote sensing images,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , 2024. 8, 9

    2024

  40. [48]

    Global sparsity-weighted local contrast measure for infrared small target detection,

    Z. Qiu, Y . Ma, F. Fan, J. Huang, and L. Wu, “Global sparsity-weighted local contrast measure for infrared small target detection,” IEEE Geoscience and Remote Sensing Letters, vol. 19, pp. 1–5, 2022. 9

    2022

  41. [49]

    Infrared small target detection based on partial sum of the tensor nuclear norm,

    L. Zhang and Z. Peng, “Infrared small target detection based on partial sum of the tensor nuclear norm,” Remote Sensing, vol. 11, no. 4, p. 382, 2019. 9

    2019

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.