ISTD-YOLO: A Multi-Scale Lightweight High-Performance Infrared Small Target Detection Algorithm

Huanbin Zhang; Ruoyan Xiong; Shang Zhang; Yujie Cui

REVIEW 4 major objections 5 minor 23 references

ISTD-YOLO: A Multi-Scale Lightweight High-Performance Infrared Small Target Detection Algorithm

T0 review · 4 major / 5 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 claims that a three-scale lightweight reconstruction of YOLOv7, with SimAM attention and an NWD-blended loss, lifts infrared small-target detection mAP0.5 from 69.3% to 77.8% on HIT-UAV and from 68.1% to 79.5% on IRSTD-1k while…

desk verdict A workmanlike YOLOv7 variant for infrared small targets whose HIT-UAV ablation is plausible, but the IRSTD-1k comparison table is mislabeled and the paper's 'all indicators improved' claim does not survive contact with its own numbers. read the letter →

arxiv 2504.14289 v1 pith:BPS4MYJX submitted 2025-04-19 cs.CV

Shang Zhang , Yujie Cui , Ruoyan Xiong , Huanbin Zhang This is my paper · ORCID

classification cs.CV

keywords infraredsmalltargetdetectionYOLOv7lightweightnetworkSimAMattentionNormalizedWassersteinDistancemulti-scaleHIT-UAVIRSTD-1k

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 ISTD-YOLO, a redesigned YOLOv7 for infrared small target detection, and claims it outperforms the original YOLOv7 and current mainstream detectors on two public infrared datasets while being considerably lighter. On HIT-UAV, mAP0.5 rises from 69.32% to 77.84% and model size falls from 71.3 MB to 21.8 MB; on IRSTD-1k, mAP0.5 rises from 68.1% to 79.5%. The gains are attributed to a three-scale architecture that swaps the 20×20 prediction head for a 160×160 head, a parameter-free SimAM attention in the neck, and an NWD-based regression loss that remains informative when boxes do not overlap. Real-time, low-cost infrared detection matters for drones and surveillance, where targets are tiny and compute is constrained.

What carries the argument

The load-bearing combination is: (1) a three-scale detection head at 1/16, 1/8 and 1/4 resolution, replacing the 20×20 head so that shallow small-target features are not down-sampled away; (2) SimAM, a parameter-free attention module that assigns each neuron a weight inferred from an energy function, inserted after the first two neck convolutions; (3) VoV-GSCSP, a one-shot aggregation module using GSConv hybrid convolution, replacing ELAN-W in the neck to cut computation; and (4) Normalized Gaussian Wasserstein Distance (NWD), a similarity measure that models each bounding box as a 2D Gaussian and compares boxes through their second-order Wasserstein distance, normalized exponentially with a dataset-related constant C. NWD is blended with IoU in the regression loss through an iou_ratio weight. The paper argues that these four pieces jointly produce the accuracy gain while shrinking the model.

What would settle it

On the HIT-UAV test split, retrain ISTD-YOLO with iou_ratio set to 0 in Eq. (8) (reverting the regression loss to plain CIoU) while keeping everything else identical; the paper's ablation row L predicts mAP0.5 should fall from 77.84% to about 75.6%. If the drop does not reproduce, or if mAP0.5 stays above 77%, the claimed contribution of the NWD term is not real. Separately, report the value of C used in Eq. (5); if C was chosen by searching on the test split, the generalization claim fails by construction.

Watch

Extended reading notes

Core claim

The central claim is that infrared small-target detection can be materially improved by rebalancing a one-stage detector toward shallow high-resolution features rather than by adding capacity. ISTD-YOLO deletes the 20×20-pixel large-receptive-field prediction head of YOLOv7 and adds a 160×160 head, so the network predicts at 1/16, 1/8 and 1/4 of input resolution; this three-scale lightweight design shifts weight onto the detail-rich early feature maps where dim small targets live. In the neck, the expensive ELAN-W blocks are replaced by the single-shot aggregation VoV-GSCSP built on GSConv, and the parameter-free attention module SimAM is inserted after the first two CBS blocks; this combination cuts parameters from 37.2M to 11.2M and GFLOPs from 105.1 to 86.6. Finally, the bounding-box regression loss blends CIoU with Normalized Gaussian Wasserstein Distance, which keeps providing useful gradients even when predicted and ground-truth boxes have zero overlap, a regime where IoU is uninformative for small targets. Together these changes are claimed to raise mAP0.5 from 69.32% to 77.84% on HIT-UAV and from 68.1% to 79.5% on IRSTD-1k. The authors explicitly note that the lightweight design may trade feature richness under extremely complex conditions, that SimAM's cross-scale adaptability needs further study, and that NWD's behaviour under heavy target overlap remains to be improved.

Load-bearing premise

The assumption that the reported accuracy gains come from the architecture itself, rather than from tuning the unreported dataset-dependent constant C in Eq. (5) and the iou_ratio weight in Eq. (8) on the test sets.

Editorial extensions

If this is right

On HIT-UAV, ISTD-YOLO outperforms YOLOv5s, YOLOv6s, YOLOv7s, YOLOv7-tiny, YOLOv8s and YOLOv8n in mAP0.5 (77.84%) with a smaller model than YOLOv7s (21.8 MB vs 71.3 MB) and an inference time of 11.5 ms.
The same configuration transfers to a second dataset, IRSTD-1k, where mAP0.5 improves from 68.1% (YOLOv7s) to 79.5%, supporting the generalization claim.
Reconfiguring the backbone alone (removing the 20×20 head, adding 160×160) cuts backbone parameters to about 45% of the original and beats both ShuffleNetV2 and MobileNetV3-Large backbones in mAP0.5 (71.9% vs 69.8% and 71.7%).
Ablation results attribute the full-model gain to the combination of components: reconfiguration alone adds 2.60 mAP0.5 points over the baseline, SimAM alone adds 3.55, LTSN adds 2.89, NWD adds 2.77, and the full combination reaches 77.84%.
If the reported comparisons hold, the recipe of high-resolution small receptive-field heads plus an overlap-insensitive loss offers a practical template for resource-constrained infrared detection systems.

Reading between the lines

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

The reported improvements are sensitive to two unreported hyperparameters: the constant C in the NWD normalization in Eq. (5) and the iou_ratio in Eq. (8). Because neither value is given, a reader cannot tell whether the reported mAP0.5 figures would survive on a new dataset without re-tuning these constants; a fair test would fix both values ahead of time.
The method is presented as specifically tuned for infrared dim-small targets, but the same recipe—small receptive-field heads, parameter-free attention, and an overlap-insensitive loss—should transfer to other small-object detection settings such as aerial imagery or medical imaging; that transfer is not demonstrated in the paper.
The model is 21.8 MB and 86.6 GFLOPs, which makes it a plausible candidate for edge deployment, but the paper reports no measured inference speed on an embedded device; the 11.5 ms figure comes from an RTX 3090, so edge-frame-rate claims remain untested.
The paper's own final paragraph concedes that the lightweight design may lose feature richness under extreme clutter and that NWD needs improvement for heavy target overlap; those regimes, such as crowded scenes or stacked targets, are exactly where the method's advantage would matter most and are not evaluated.

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.

Desk Editor's Note

read the letter

Short version: this is a routine combination of known components—network reconfiguration, SimAM, GSConv/VoV-GSCSP, and NWD—applied to infrared small target detection. The HIT-UAV ablation is systematically presented and the improvement over YOLOv7s (69.32 to 77.84 mAP0.5) is real if you trust the training setup. But the generalization evidence is shaky, and the stress-test note about Table 6 is accurate.

First, one correction to the reader's note: row L of Table 3 does include LTSN. The parameter counts in the ablation are actually consistent—adding LTSN drops the model from 23.8 to 21.8 MB, and SimAM/NWD add no parameters. So that particular complaint is a misreading.

The real problem is Table 6. The YOLOv5s and YOLOv8s rows list 2.5M and 3.0M parameters. Those are not the -s variants; they match the -n variants (YOLOv8n is ~3M params, YOLOv5n is close to 2.5M at some input size). The same paper's Table 5 reports YOLOv8s with a 21.4 MB model, which is the real -s scale. So the table is internally inconsistent. If the baselines are actually the n variants, then ISTD-YOLO beats them by only 0.2–0.3 mAP points, has lower recall than YOLOv8s (75.5 vs 76.0), while using 3–4x parameters and ~10x FLOPs. That directly contradicts the abstract's 'all indicators are effectively improved.'

Additional soft spots: the NWD constant C in Eq. (5) is 'closely related to the data set' but never specified, and iou_ratio in Eq. (8) is also unreported. Without these, the ablation gains attributed to NWD could be tuned on the test set. No code, no random seeds, no error bars. Also, on HIT-UAV the inference time is 11.5 ms vs YOLOv8s at 5.7 ms, so 'lightweight' is relative to YOLOv7s, not to modern efficient detectors.

What the paper does well: it is honest about combining prior work, the limitations paragraph is more candid than most papers in this genre, and the ablation design covers the main components. The backbone change (removing 20x20, adding 160x160) is a sensible specialization for small targets.

Who is this for? A reader who wants a data point on YOLOv7 variants for infrared small target detection. It is not a novel mechanism paper. I would not cite it in my own work. It deserves peer review only if the authors can correct Table 6, report the constants, and ideally release code. A serious editor could also desk reject it as incremental with flawed supporting evidence. My recommendation: send it to review with a strong request for revision, but treat the central generalization claim as unproven until Table 6 is fixed and the hyperparameters are disclosed.

Referee Report

4 major / 5 minor

Summary. The paper proposes ISTD-YOLO, an infrared small-target detector built by reconfiguring YOLOv7. It removes the 20×20 prediction scale, adds a 160×160 scale, uses a lightweight neck (LTSN) based on GSConv/VoV-GSCSP, inserts a parameter-free SimAM attention module, and replaces part of the box-regression loss with a normalized Wasserstein distance (NWD). Ablations on HIT-UAV show mAP0.5 increasing from 69.32% for YOLOv7s to 77.84%, with model size reduced from 71.3 MB to 21.8 MB; a generalization test on IRSTD-1k reports mAP0.5 of 79.5% versus 68.1% for YOLOv7s. The authors conclude that ISTD-YOLO outperforms YOLOv7 and current mainstream algorithms on all indicators.

Significance. If the results are reproducible, this is a useful practical engineering contribution: it combines known components (three-scale prediction, SimAM, GSConv/VoV-GSCSP, NWD) into a configuration that appears to improve HIT-UAV performance at lower complexity. The modular ablations in Table 3 are systematic, and the model-size pattern is internally consistent: rows containing LTSN show 21.8 MB/11.2 M parameters, while rows without LTSN show 23.8–23.9 MB/12.3 M. However, the paper does not provide code, random seeds, or training hyperparameters, and the generalization comparison in Table 6 has an internal inconsistency that touches the paper's central claim. The contribution is therefore currently conditional on correcting these issues and on presenting the missing configuration details.

major comments (4)

[§4.7, Table 6] The YOLOv5s and YOLOv8s baselines are labeled with parameter counts 2.5 M and 3.0 M and GFLOPs 7.2 and 8.2, which are the standard values of the nano variants (YOLOv5n and YOLOv8n), not the small variants. Moreover, Table 5 in the same paper reports YOLOv8s as a 21.4 MB model, so the YOLOv8s in Table 6 cannot be the same model. If the tabulated baselines are actually nano variants, then the reported mAP margins of 0.2–0.3 points, with lower recall than YOLOv8s (75.5 vs 76.0), do not support the abstract's claim that 'all indicators are effectively improved' relative to mainstream algorithms; if the baselines are genuine small variants, the parameter counts are wrong. This inconsistency must be resolved by correcting the table or rerunning the comparison.
[§3.4, Eqs. (5) and (8)] The NWD loss depends on the constant C, which the text states is 'closely related to the data set,' and on the balancing coefficient iou_ratio, but neither value is reported for either dataset. Because rows D, I, J, L, M, and ISTD-YOLO in Table 3 attribute improvement to NWD, the missing values make the ablation non-reproducible and leave open the possibility that part of the gain comes from tuning these constants rather than from the architectural change. Please report C and iou_ratio for HIT-UAV and IRSTD-1k and, if they were chosen on validation data, describe the selection procedure.
[§4.3 and §4.7, Tables 3 and 6] All results are reported for a single run without random seeds, error bars, or repeated trials. On IRSTD-1k the difference between ISTD-YOLO and the tabulated YOLOv8s baseline is only 0.3 mAP0.5 (79.5 vs 79.2) with lower recall (75.5 vs 76.0); without variance estimates this margin is indistinguishable from run-to-run noise. The claim that all indicators improve over mainstream algorithms is therefore not statistically supported, and the same limitation applies to the smaller ablation gaps in Table 3.
[§3.1, §3.3, and Table 5] The text repeatedly emphasizes that the lightweight design reduces computation and speeds up inference, but the reported numbers are more nuanced: Table 3 gives GFLOPs of 99.1 for the reconfiguration alone versus 105.1 for YOLOv7s, and Table 5 shows ISTD-YOLO inference time of 11.5 ms, which is slower than YOLOv8s (5.7 ms) and YOLOv8n (4.9 ms) on HIT-UAV. The claims of computational advantage should be qualified to the YOLOv7 baseline and should not be extended to 'current mainstream algorithms' without a more complete comparison.

minor comments (5)

[Figure 8] The caption labels both the second and third columns as '(b)'; the third column should be labeled '(c)'.
[Eq. (1)] Equation (1) and the surrounding definitions of mu and sigma are typeset in a garbled way; the energy function and the variable definitions should be rewritten for readability.
[§3.5] The sentence 'the feature map output of 20×20 pixel scale was removed at the preresetting end' appears to contain a typo; 'preresetting' should likely be 'prediction'.
[Table 2] Training details such as optimizer, initial learning rate, momentum, weight decay, and data augmentation are not reported; only batch size, epochs, and input size are given, which is insufficient for reproduction.
[§4.5] The confusion-matrix discussion reports increases of '6% and 20%' for OtherVehicle and DontCare, but the figure is not legible enough to verify these numbers; please ensure the confusion matrices are readable and that the percentages are precisely stated.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is an empirical architecture comparison against external baselines; the unreported NWD constants are a reproducibility gap, not a circular reduction.

full rationale

ISTD-YOLO is an engineering paper whose claims are empirical comparisons on public datasets (HIT-UAV, IRSTD-1k) against external baselines. There is no derivation chain in which a predicted quantity is defined in terms of the target result. The backbone reconfiguration, SimAM insertion, VoV-GSCSP/GSConv neck, and NWD loss are all taken from cited external work and combined, and their contributions are evaluated by measured mAP, precision, recall, parameters, and FLOPs. The central claim that ISTD-YOLO improves over YOLOv7 and other YOLO variants is not equivalent to its inputs: the architecture does not encode the reported mAP values and the tables compare against independent baselines. The only weakness close to circularity is that Eq. (5) defines NWD with a constant C that the text states is closely related to the data set, and Eq. (8) introduces an unreported iou_ratio balancing NWD and IoU; if these were tuned on the test sets, the ablation gains attributed to NWD would not establish a general property. That is a missing-support/reproducibility concern, not a self-referential reduction by construction, and no fitted parameter is relabeled as a prediction. Similarly, the possible inconsistency in Table 6's baseline labels or parameter counts is a correctness/evidence issue, not circularity. Accordingly, the appropriate finding is no significant circularity.

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

The paper's central claim rests on standard YOLOv7 machinery plus three externally published components. The only entities introduced by hand are the NWD constant C and the loss-balance weight iou_ratio, both unreported, and the feature scale configuration. No new physical or architectural entity is invented.

free parameters (3)

NWD constant C = not reported
Eq. (5) defines NWD = exp(-W2^2/C) with C 'a constant closely related to the data set'; its value is never given, so it is a tuning parameter for the loss.
iou_ratio in box loss = not reported
Eq. (8) mixes NWD and IoU losses using iou_ratio; the chosen ratio is not stated, allowing the loss balance to be tuned per dataset.
output feature map scales = 40x40, 80x80, 160x160
The choice to replace the 20x20 head with a 160x160 head is made by hand to favor small targets; it is a design choice not derived from data.

assumptions (4)

domain assumption SimAM energy function and its parameter-free property
The paper adopts SimAM from [14] and assumes it provides useful 3D attention weights without adding parameters; Eq. (1) is reproduced from the cited work.
domain assumption GSConv and VoV-GSCSP reduce cost without hurting accuracy
The lightweight neck relies on GSConv/VoV-GSCSP from [18]; the paper assumes their published efficiency-accuracy trade-off transfers to infrared small-target detection.
domain assumption NWD loss improves small-target localization
The normalized Wasserstein distance is taken from [19] and prior NWD literature; the paper assumes it is scale-insensitive and better than IoU for tiny boxes.
domain assumption Evaluation metrics and dataset splits are representative
The paper assumes mAP0.5, P, and R on HIT-UAV and IRSTD-1k with a 70/20/10 or 70/15/15 split are sufficient to demonstrate generalization; no statistical tests or multiple seeds are provided.

how reviews work

0 comments

Cite this review

Pith. "Pith review of ISTD-YOLO: A Multi-Scale Lightweight High-Performance Infrared Small Target Detection Algorithm." pith.science (2026). https://pith.science/paper/BPS4MYJX

@misc{pith2026250414289,
  author       = {Pith},
  title        = {Pith review of: ISTD-YOLO: A Multi-Scale Lightweight High-Performance Infrared Small Target Detection Algorithm},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BPS4MYJX}},
  note         = {Machine review of arXiv:2504.14289}
}

read the original abstract

Aiming at the detection difficulties of infrared images such as complex background, low signal-to-noise ratio, small target size and weak brightness, a lightweight infrared small target detection algorithm ISTD-YOLO based on improved YOLOv7 was proposed. Firstly, the YOLOv7 network structure was lightweight reconstructed, and a three-scale lightweight network architecture was designed. Then, the ELAN-W module of the model neck network is replaced by VoV-GSCSP to reduce the computational cost and the complexity of the network structure. Secondly, a parameter-free attention mechanism was introduced into the neck network to enhance the relevance of local con-text information. Finally, the Normalized Wasserstein Distance (NWD) was used to optimize the commonly used IoU index to enhance the localization and detection accuracy of small targets. Experimental results show that compared with YOLOv7 and the current mainstream algorithms, ISTD-YOLO can effectively improve the detection effect, and all indicators are effectively improved, which can achieve high-quality detection of infrared small targets.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 22 canonical work pages

[1]

IEEE Transactions on Geoscience and Remote Sensing, 56(4), 2452-2466 (2018)

Bai, X., Bi, Y.: Derivative Entropy-Based Contrast Measure for Infrared Small-Target De- tection. IEEE Transactions on Geoscience and Remote Sensing, 56(4), 2452-2466 (2018)

work page 2018
[2]

In: IEEE International Conference on Computer Vision, 1440- 1448 (2015)

Girshick, R.: Fast R-CNN. In: IEEE International Conference on Computer Vision, 1440- 1448 (2015)

work page 2015
[3]

IEEE Transactions on Pattern Analysis and Machine Intel- ligence, 39(6), 1137-1149 (2017)

Ren, S., He, K., Girshick, R., Sun, J.: Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. IEEE Transactions on Pattern Analysis and Machine Intel- ligence, 39(6), 1137-1149 (2017)

work page 2017
[4]

In: IEEE International Confer- ence on Computer Vision, 2980-2988 (2017)

He, K., Gkioxari, G., Dollár, P., Girshick, R.: Mask R-CNN. In: IEEE International Confer- ence on Computer Vision, 2980-2988 (2017)

work page 2017
[5]

IEEE Transactions on Geoscience and Remote Sensing, 60, 1-14 (2022)

Li, S., Cui, X., Li, G., Zhang, L., Chen, X., Cao, X.: Enhanced Automatic Root Recognition and Localization in GPR Images Through a YOLOv4-Based Deep Learning Approach . IEEE Transactions on Geoscience and Remote Sensing, 60, 1-14 (2022)

work page 2022
[6]

T., Lee, H

Kim, M., Oh, K., Cho, Y., Seo, H., Nguyen, X. T., Lee, H. -J.: A Low-Latency FPGA Ac- celerator for YOLOv3-Tiny With Flexible Layerwise Mapping and Dataflow. IEEE Trans- actions on Circuits and Systems I: Regular Papers, 71(3), 1158-1171 (2024)

work page 2024
[7]

In: IEEE Conference on Computer Vision and Pattern Recognition, 6517-6525 (2017)

Redmon, J., Farhadi, Ali.: YOLO9000: Better, Fast er, Stronger. In: IEEE Conference on Computer Vision and Pattern Recognition, 6517-6525 (2017)

work page 2017
[8]

IEEE Internet of Things Journal, 11(20), 32804-32812 (2024)

Zhao, H., Gao, Y., Deng, W.: Defect Detection Using Shuffle Net -CA-SSD Lightweight Network for Turbine Blades in IoT. IEEE Internet of Things Journal, 11(20), 32804-32812 (2024)

work page 2024

Show all 23 references

[9]

In: IEEE/CVF International Conference on Computer Vision, 6568-6577 (2019) 18 S

Duan, K., Bai, S., Xie, L., Qi, H., Huang, Q., Tian, Q.: CenterNet: Keypoint Triplets for Object Detection. In: IEEE/CVF International Conference on Computer Vision, 6568-6577 (2019) 18 S. Zhang et al

2019
[10]

IEEE Transactions on Pattern Analysis and Machine Intelligence, 46(5), 3509-3521 (2024)

Duan, K., Bai, S., Xie, L., Qi, H., Huang, Q., Tian, Q.: CenterNet++ for Object Detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, 46(5), 3509-3521 (2024)

2024
[11]

H., Park, C

Lee, I. H., Park, C. G.: Infrared Small Target Detection Algorithm Using an Augmented Intensity and Density-Based Clustering . IEEE Transactions on Geoscience and Remote Sensing, 61, 1-14 (2023)

2023
[12]

Journal of Computational Methods in Sci- ences and Engineering, 22(4), 1393-1408 (2022)

Liu, G., Cao, Z., Liu , S., Song, B., Liu, Z.: An improved SSD method for infrared target detection based on convolutional neural network. Journal of Computational Methods in Sci- ences and Engineering, 22(4), 1393-1408 (2022)

2022
[13]

L., Lefebvr, S., Woiselle, A., Barbanson, C.: A Contrario Paradigm for Yolo-Based Infrared Small Target Detection

Ciocarlan, A., Hegarat-Mascle, S. L., Lefebvr, S., Woiselle, A., Barbanson, C.: A Contrario Paradigm for Yolo-Based Infrared Small Target Detection . In: IEEE International Confer- ence on Acoustics, Speech and Signal Processing, 5630-5634 (2024)

2024
[14]

In : International Conference on Machine Learning , (2021)

Yang, L., Zhang, R., Li, L., Xie, X.: SimAM: A Simple, Parameter-Free Attention Module for Convolutional Neural Networks . In : International Conference on Machine Learning , (2021)

2021
[15]

IEEE Internet of Things Journal, 12(5), 5822-5833 (2025)

Jin, J., Wu, X., Daly, I., Chen, W., He, X., Wang, X.: Squeeze and Excitation-Based Mul- tiscale CNN for Classification of Steady -State Visual Evoked Potentials . IEEE Internet of Things Journal, 12(5), 5822-5833 (2025)

2025
[16]

IEEE International Conference on Communications, 4066-4071 (2024)

Deng, X., Ma, Y., Zhang, X., Zhu, H.: Convolutional Block Attention Module-Based Neural Network for Enhanced IQ Imbalance Estimation in Low Signal -to-Noise Ratio Environ- ments. IEEE International Conference on Communications, 4066-4071 (2024)

2024
[17]

IEEE Transactions on Neural Networks and Learning Systems, 35(7), 8984-8998 (2024)

Tian, Y., Shen, Y., Wang, X., Wang, J., Wang, K., Ding, W., Wang, Z., Wang, F.: Learning Lightweight Dynamic Kernels With Attention Inside via Local –Global Context Fusion . IEEE Transactions on Neural Networks and Learning Systems, 35(7), 8984-8998 (2024)

2024
[18]

arXiv preprint arXiv:2206.02424 (2024)

Li, H., Li, J., Wei, H., Li, Z., Zhan, Z., Ren, Q.: Slim-neck by GSConv: A lightweight-design for real-time detector architectures. arXiv preprint arXiv:2206.02424 (2024)

2024 arXiv
[19]

IEEE Transactions on Geoscience and Remote Sensing, 62, 1-11 (2024)

Yang, B., Zhang, X., Zhang, J., Luo, J., Zhou, M., Pi, Y.: EFLNet: Enhancing Feature Learn- ing Network for Infrared Small Target Detection . IEEE Transactions on Geoscience and Remote Sensing, 62, 1-11 (2024)

2024
[20]

Scientific Data, 10, 227 (2023)

Suo, J., Wang, T., Zhang, X., Chen, H., Zhou, W., Shi, W.: HIT-UAV: A high-altitude in- frared thermal dataset for Unmanned Aerial Vehicle-based object detection. Scientific Data, 10, 227 (2023)

2023
[21]

In IEEE/CVF Conference on Computer Vision and Pattern Recognition, 867–876 (2022)

Zhang, M., Zhang, R., Yang, Y., Bai, H., Zhang, J., Guo, J.: ISNet: Shape Matters for Infra- red Small Target Detection . In IEEE/CVF Conference on Computer Vision and Pattern Recognition, 867–876 (2022)

2022
[22]

In: European Conference on Computer Vision, 122-138 (2018)

Ma, N., Zhang, X., Zheng, HT., Sun, J.: ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design. In: European Conference on Computer Vision, 122-138 (2018)

2018
[23]

In IEEE/CVF Interna- tional Conference on Computer Vision, 1314-1324 (2020)

Howard, A., Sandler, M., Chen, B., Wang, W., Chen, L., Tan, M., Chu, G., Vasudevan, V., Zhu, Y., Pang, R., Adam, H., Le , Q.: Searching for MobileNetV3 . In IEEE/CVF Interna- tional Conference on Computer Vision, 1314-1324 (2020)

2020

Pith tools

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

[1] [1]

IEEE Transactions on Geoscience and Remote Sensing, 56(4), 2452-2466 (2018)

Bai, X., Bi, Y.: Derivative Entropy-Based Contrast Measure for Infrared Small-Target De- tection. IEEE Transactions on Geoscience and Remote Sensing, 56(4), 2452-2466 (2018)

work page 2018

[2] [2]

In: IEEE International Conference on Computer Vision, 1440- 1448 (2015)

Girshick, R.: Fast R-CNN. In: IEEE International Conference on Computer Vision, 1440- 1448 (2015)

work page 2015

[3] [3]

IEEE Transactions on Pattern Analysis and Machine Intel- ligence, 39(6), 1137-1149 (2017)

Ren, S., He, K., Girshick, R., Sun, J.: Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. IEEE Transactions on Pattern Analysis and Machine Intel- ligence, 39(6), 1137-1149 (2017)

work page 2017

[4] [4]

In: IEEE International Confer- ence on Computer Vision, 2980-2988 (2017)

He, K., Gkioxari, G., Dollár, P., Girshick, R.: Mask R-CNN. In: IEEE International Confer- ence on Computer Vision, 2980-2988 (2017)

work page 2017

[5] [5]

IEEE Transactions on Geoscience and Remote Sensing, 60, 1-14 (2022)

Li, S., Cui, X., Li, G., Zhang, L., Chen, X., Cao, X.: Enhanced Automatic Root Recognition and Localization in GPR Images Through a YOLOv4-Based Deep Learning Approach . IEEE Transactions on Geoscience and Remote Sensing, 60, 1-14 (2022)

work page 2022

[6] [6]

T., Lee, H

Kim, M., Oh, K., Cho, Y., Seo, H., Nguyen, X. T., Lee, H. -J.: A Low-Latency FPGA Ac- celerator for YOLOv3-Tiny With Flexible Layerwise Mapping and Dataflow. IEEE Trans- actions on Circuits and Systems I: Regular Papers, 71(3), 1158-1171 (2024)

work page 2024

[7] [7]

In: IEEE Conference on Computer Vision and Pattern Recognition, 6517-6525 (2017)

Redmon, J., Farhadi, Ali.: YOLO9000: Better, Fast er, Stronger. In: IEEE Conference on Computer Vision and Pattern Recognition, 6517-6525 (2017)

work page 2017

[8] [8]

IEEE Internet of Things Journal, 11(20), 32804-32812 (2024)

Zhao, H., Gao, Y., Deng, W.: Defect Detection Using Shuffle Net -CA-SSD Lightweight Network for Turbine Blades in IoT. IEEE Internet of Things Journal, 11(20), 32804-32812 (2024)

work page 2024