Pith. sign in

REVIEW 3 major objections 1 minor 35 references

MSFNet-CPD: Multi-Scale Cross-Modal Fusion Network for Crop Pest Detection

T0 review · 3 major / 1 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 A pest detector that fuses super-resolved images with species text descriptions claims to beat state-of-the-art unimodal detectors on IP102-derived benchmarks.

desk verdict The multimodal pest-detection results are likely inflated by label leakage through the per-species text input; the datasets and architecture are new but the main empirical claim isn't supported by the reported protocol. read the letter →

arxiv 2505.02441 v1 pith:KZB36Y4W submitted 2025-05-05 cs.AI

Jiaqi Zhang , Zhuodong Liu , Kejian Yu This is my paper · ORCID
classification cs.AI
keywords pestdetectioncross-modalfusionimage-textsuper-resolutionIP102datasetobjectmulti-modaldataaugmentation
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 claims that agricultural pest detection can be improved by giving the detector two extra information channels beyond the raw image: a super-resolution reconstruction of the same image, and a short text description of the target species' appearance. The proposed MSFNet-CPD feeds both the original and the reconstructed image through a multi-scale convolutional stem, converts the visual features into token sequences, and mixes them with text tokens in a Transformer-style fusion module before the YOLOv4-style detection head. On the paper's new benchmarks built from IP102, the model reports precision 82.15% and mAP 46.06% on CTIP102, outperforming the strongest unimodal baseline YOLOv9 across mAP, mAP50 (92.18% vs 81.10%), and mAP75. Ablations show that removing the text channel costs about 10 points of mAP and removing the super-resolution path costs about 11 points, so both channels are load-bearing in the reported result.

What carries the argument

The load-bearing mechanism is the ITF module: a Transformer encoder (scaled dot-product attention, multi-head self-attention plus feed-forward sub-layers) that receives a concatenated sequence G built from n text tokens and 25 visual tokens at each of three scales, for both the original and super-resolved image, so the same attention operation mixes semantic class cues with fine-grained visual cues. The TIC and ITC converters map visual feature maps into and out of this token space; the LSRGAN super-resolution path supplies a second, sharper view of the same scene; and the YOLOv4 neck/head (SPP + PANet) turns the fused multi-scale features into bounding-box predictions.

What would settle it

Run MSFNet-CPD on CTIP102 with the text channel removed, or with text descriptions swapped randomly between species; the paper's own ablation predicts mAP falls from 46.06% to about 35.6%, which would erase most of the claimed advantage over YOLOv9 (42.32%) on the same data.

Watch

Extended reading notes

Core claim

The central discovery the authors report is that cross-modal fusion of visual and textual features beats unimodal detection on pest benchmarks, provided the text describes the species present. The network architecture interleaves three components: LSRGAN super-resolution to repair low-quality field images; a Text-Image Converter (TIC) that condenses multi-scale visual features into 25 tokens per scale; and an Image-Text Fusion (ITF) Transformer in which text tokens and visual tokens attend to one another, with an Image-Text Converter (ITC) decoding the fused representation back to feature maps for the detection neck. The paper also contributes three datasets—STIP102 (simple text), CTIP102 (complex text), and MTIP102 (multi-target composites made by the ACIE augmentation algorithm)—and reports that more complex text yields higher precision and mAP50 than simple text.

Load-bearing premise

The evaluation assumes it is a valid protocol to hand the model a text description built from the species label at inference time; if that text is removed or does not match the image's species, the reported gains largely disappear.

Editorial extensions

If this is right

  • If text descriptions of the target species are available at inference time, pest detectors can be expected to beat image-only detectors by a wide margin on IP102-style data.
  • The roughly 10-point mAP drop without text indicates that almost all of the multimodal advantage depends on the semantic channel, not on the visual architecture alone.
  • The super-resolution path contributes about as much as the text channel, suggesting low-quality field images are a first-order bottleneck for pest detection.
  • The ACIE augmentation can generate labeled multi-target training images with random backgrounds and placements, which the paper reports improves robustness in complex scenes even if raw scores on MTIP102 are lower.

Reading between the lines

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

  • Because the text descriptions in STIP102 and CTIP102 are constructed from the species identities, the reported mAP measures how much class-label information the model can exploit through the text channel; a deployment protocol with genuinely independent text (e.g., a field-guide entry not tailored to the image) would likely show smaller gains.
  • The same TIC-ITF-ITC design could transfer to any fine-grained detection task—plant diseases, medical images, insect taxonomy—where a short verbal description of the target class exists, offering a cheap way to inject semantic priors.
  • A fairer comparison against unimodal baselines would either give those baselines equivalent class information or evaluate MSFNet-CPD under text-only, no-text, and mismatched-text conditions; the paper reports only the no-text ablation.
  • The roughly 10-point gap without text suggests the visual backbone alone may underperform YOLOv9, so the headline comparison effectively contrasts a text-augmented model with an image-only model.
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

3 major / 1 minor

Summary. The paper proposes MSFNet-CPD, a multi-scale cross-modal fusion network for crop pest detection. The method combines super-resolved and original images with per-species text descriptions, and the authors introduce four datasets (HIP102, STIP102, CTIP102, MTIP102) derived from IP102, plus an ACIE data-enhancement algorithm for multi-target images. The central claim is that MSFNet-CPD consistently outperforms state-of-the-art unimodal detectors on these benchmarks, with the best result reported on the CTIP102 dataset (mAP 46.06, mAP50 92.18, mAP75 40.18).

Significance. If the evaluation protocol were sound, the paper would make a useful contribution by introducing multimodal pest-detection benchmarks and demonstrating a cross-modal benefit. The datasets, particularly STIP102 and CTIP102, could be of interest to the community, and the ACIE algorithm addresses a practical multi-target scenario. However, the central claim of cross-modal superiority rests on an evaluation setup in which the test-time text input is generated from the ground-truth species label. This confound, together with uncontrolled comparisons against published baselines, means the reported gains are not evidence for the paper's central conclusion. The strength of the contribution therefore depends entirely on whether the protocol can be fixed, which the current manuscript does not demonstrate.

major comments (3)
  1. [§II-F, Figure 3; §III-C; Table VI] The comparison in Table IV versus Table V is not controlled. Table IV reports YOLOv9 on unimodal IP102 at 42.32 mAP, while Table V reports MSFNet-CPD on CTIP102 at 46.06 mAP. These differ in the dataset (IP102 vs CTIP102), in the training data (the multimodal model also sees text, which leaks the label), and possibly in the train/validation/test split: Section III-A states only that datasets are divided 'according to 8:1:1' with no random seed or confirmation that the split matches that used for the published IP102 baselines in Table IV. No experiment runs MSFNet-CPD on unmodified IP102 under the same training conditions as the baselines, and no error bars or repeated runs are reported. The 0.32-point mAP75 difference (Table V 40.18 vs Table IV 39.86) is well within run-to-run noise for object detectors, so the claim of 'significantly outperforms this state-of-the-art model across all metrics' is not supported by the presented evidence.
  2. [Table V vs Table VI] There is an internal inconsistency in the reported mAP75 values for the full model on CTIP102. Table V lists the CTIP102 row as P=82.15, F1=78.21, mAP=46.06, mAP50=92.18, mAP75=40.18, while Table VI lists 'Our model' with mAP=46.06, mAP50=92.18, mAP75=39.86. The same setup should produce the same number; if the difference is due to a different dataset split or random seed, that should be stated. As written, this discrepancy undermines the reliability of the numerical comparisons, especially since the mAP75 value in Table VI coincides exactly with the YOLOv9 mAP75 in Table IV.
  3. [§III-A and Table II] The ACIE-generated MTIP102 results are not compared against any baseline detector on the same data. Table V reports only MSFNet-CPD on MTIP102 (mAP 22.33), and Section III-E attributes the drop to task complexity. Without a unimodal baseline trained on MTIP102, the reader cannot tell whether the low score is due to the difficulty of the multi-target task or to a deficiency of the proposed method on that data.
minor comments (1)
  1. [§III-C] Section II-G states parameter values for ACIE (B=580, T=820, R=4, num=10000) without any justification or sensitivity analysis; the choice of these values may materially affect the MTIP102 results.

Circularity Check

1 steps flagged · score 8.0 of 10

Text input is built from ground-truth species labels; the reported cross-modal gain is a label side-channel by construction.

  1. self definitional [Section II-F (Data Set Construction), Figure 3; Section III-C; Table VI]
    "we constructed two text descriptions, simple and complex, for each pest. These text descriptions are intended to provide information about the pest characteristics ... The results indicate that the CTIP02 and STIP102 datasets outperform the unimodal dataset in detection performance. This suggests that incorporating textual descriptive features corresponding to the images helps overcome the limitations of using single-image features. ... Simple Description: The adult body of the rice leaf roller is approximately 12-15 millimeters long and appears grayish brown."

    The text descriptions are authored per pest species (Figure 3), and at test time each image is paired with the description of its true species. Thus the text modality carries the ground-truth class identity; the network can read the answer class from the text. The claimed cross-modal advantage over unimodal IP102 baselines (e.g., CTIP102 46.06 mAP vs YOLOv9 42.32) is not an independent test: the baselines never receive the text side channel. The paper's own ablation confirms the channel's role: removing text lowers mAP by 10.42 points, roughly the reported margin. The conclusion that cross-modal fusion is more effective than unimodal features therefore reduces by construction to feeding class-conditioned text into the classifier.

full rationale

The central empirical claim is that MSFNet-CPD consistently outperforms state-of-the-art unimodal detectors on pest detection benchmarks, and that this superiority demonstrates the benefit of cross-modal fusion. The evaluation protocol, however, supplies test images with text descriptions that are constructed from the species label of each image: Figure 3 gives per-species descriptions, and Table II pairs every image with simple or complex text. At inference the model therefore receives the target class through the text modality, while the unimodal baselines in Table IV do not. The ablation in Table VI shows that removing text drops mAP by 10.42 points, which is comparable to the gap between CTIP102 and YOLOv9, indicating that most of the claimed cross-modal gain is attributable to the class-identifying text input rather than to learned visual-text fusion. This is not a controlled test of multimodal fusion, and the abstract's 'consistently outperforms' claim is not supported as a statement about multimodal capability. The paper also contains no external benchmark validation in which text is withheld or class-agnostic; all comparisons are on datasets where the text is derived from the labels. For this reason the central claim is substantially circular by construction, though the super-resolution and data-augmentation components are evaluated separately and are not themselves circular. The score reflects that the headline cross-modal result is forced by the input protocol, not that the entire architecture is vacuous.

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

The central results rest on the new datasets and the fusion network; the numbers chosen for text lengths, ACIE parameters, and the unstated split all affect the reported metrics. The text itself, being derived from class labels, is the largest unaccounted-for input.

free parameters (5)
  • wordmaxlen = 41
    Maximum number of text tokens; truncates the per-species description and controls how much label-derived text the model sees.
  • sentmaxlen = 35
    Maximum number of sentences in the text input; chosen by hand.
  • numattentionheads = 16
    Number of attention heads in ITF; inconsistent with the 12 heads of bert-base-uncased, suggesting a custom configuration.
  • ACIE targets per image (R) = 4
    Number of pest instances pasted onto each synthetic background in MTIP102; directly sets task difficulty.
  • ACIE num generated images = 10000
    Size of the synthetic MTIP102 dataset; chosen by hand.
assumptions (4)
  • domain assumption IP102 annotations are correct and the image-label correspondence is noise-free.
    Everything downstream uses IP102 labels as ground truth; no manual verification of labels is reported.
  • domain assumption The per-species text descriptions collected from books and websites are accurate and aligned with the IP102 class labels.
    Section II-F; this is the source of the label leak because descriptions are written from the class names.
  • domain assumption Pretrained Real-ESRGAN and BERT produce features that help pest detection in this dataset.
    The model relies on these pretrained components without fine-tuning or domain adaptation analysis.
  • domain assumption The 8:1:1 random split is equivalent to the IP102 split used for published baselines.
    Section III-A; if the split differs from the standard IP102 split, the Table IV baseline numbers are not comparable.
invented entities (4)
  • STIP102
    purpose: Benchmark pairing IP102 images with simple per-species text descriptions.
    Not released; no external validation; the text is class-derived.
  • CTIP102
    purpose: Benchmark with complex per-species text descriptions.
    Not released; no external validation; the text is class-derived.
  • HIP102
    purpose: Hand-screened high-quality subset of IP102 with text.
    Selection criteria not specified; not released.
  • MTIP102
    purpose: Synthetic multi-target pest dataset generated by pasting pest crops on backgrounds.
    Generation algorithm has a documented bug in the overlap branch; not released.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MSFNet-CPD: Multi-Scale Cross-Modal Fusion Network for Crop Pest Detection." pith.science (2026). https://pith.science/paper/KZB36Y4W

@misc{pith2026250502441,
  author       = {Pith},
  title        = {Pith review of: MSFNet-CPD: Multi-Scale Cross-Modal Fusion Network for Crop Pest Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KZB36Y4W}},
  note         = {Machine review of arXiv:2505.02441}
}
read the original abstract

Accurate identification of agricultural pests is essential for crop protection but remains challenging due to the large intra-class variance and fine-grained differences among pest species. While deep learning has advanced pest detection, most existing approaches rely solely on low-level visual features and lack effective multi-modal integration, leading to limited accuracy and poor interpretability. Moreover, the scarcity of high-quality multi-modal agricultural datasets further restricts progress in this field. To address these issues, we construct two novel multi-modal benchmarks-CTIP102 and STIP102-based on the widely-used IP102 dataset, and introduce a Multi-scale Cross-Modal Fusion Network (MSFNet-CPD) for robust pest detection. Our approach enhances visual quality via a super-resolution reconstruction module, and feeds both the original and reconstructed images into the network to improve clarity and detection performance. To better exploit semantic cues, we propose an Image-Text Fusion (ITF) module for joint modeling of visual and textual features, and an Image-Text Converter (ITC) that reconstructs fine-grained details across multiple scales to handle challenging backgrounds. Furthermore, we introduce an Arbitrary Combination Image Enhancement (ACIE) strategy to generate a more complex and diverse pest detection dataset, MTIP102, improving the model's generalization to real-world scenarios. Extensive experiments demonstrate that MSFNet-CPD consistently outperforms state-of-the-art methods on multiple pest detection benchmarks. All code and datasets will be made publicly available at: https://github.com/Healer-ML/MSFNet-CPD.

Figures

Figures reproduced from arXiv: 2505.02441 by the authors.

Figure 1
Figure 1. MSFNet-CPD Model Architecture (A. LSRGAN for super-resolution of low-quality images, B. Picture to ITF [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. ITF Specific Process. (Bi , Ci , Vi stand for different scale image features and T stands for text features). TABLE I: TIC and ITC Structure. TIC B1 Conv [K=3x3, P=1, S=2] MaxPooling [2x2] B2 Conv [K=3x3, P=1, S=2; K=3x3, P=1, S=2] MaxPooling [2x2] B3 Conv [K=5x5, P=2, S=4; K=3x3, P=1, S=2] MaxPooling [3x3] ITC C1 UpSampling [2x2] ConvT [K=3x3, P=1, S=2] C2 UpSampling [2x2] ConvT [K=3x3, P=1, S=2; K=3x3, P=1, S=2] C… view at source ↗
Figure 3
Figure 3. Partial Presentation of Multi-modal Dataset. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Visualization of the semantic correlation analysis of [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: Comparison of Indicators for Different Data Sets. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Grad-cam Visualization Results. (Single modal images [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 23 canonical work pages

  1. [29]

    Cam-can: Class activation map-based categorical adversarial network,

    G. Batchuluun, J. Choi, and K. R. Park, “Cam-can: Class activation map-based categorical adversarial network,” Expert Systems with Appli- cations, vol. 222, p. 119809, 2023

    work page 2023

  2. [1]

    Pplc-net: Neural network- based plant disease identification model supported by weather data augmentation and multi-level attention mechanism,

    G. Dai, J. Fan, Z. Tian, and C. Wang, “Pplc-net: Neural network- based plant disease identification model supported by weather data augmentation and multi-level attention mechanism,” Journal of King Saud University-Computer and Information Sciences , vol. 35, no. 5, p. 101555, 2023

    work page 2023

  3. [2]

    Offline mobile diagnosis system for citrus pests and diseases using deep compression neural network,

    J. You and J. Lee, “Offline mobile diagnosis system for citrus pests and diseases using deep compression neural network,” IET Computer Vision, vol. 14, no. 6, pp. 370–377, 2020

    work page 2020

  4. [3]

    Identification of crop diseases using improved convolutional neural networks,

    L. Wang, J. Sun, X. Wu, J. Shen, B. Lu, and W. Tan, “Identification of crop diseases using improved convolutional neural networks,” IET Computer Vision, vol. 14, no. 7, pp. 538–545, 2020

    work page 2020

  5. [4]

    An efficient insect pest clas- sification using multiple convolutional neural network based models,

    H. T. Ung, H. Q. Ung, and B. T. Nguyen, “An efficient insect pest clas- sification using multiple convolutional neural network based models,” arXiv preprint arXiv:2107.12189 , 2021

    arXiv 2021

  6. [5]

    A novel multi-label pest image classifier using the modified swin transformer and soft binary cross entropy loss,

    Q. Guo, C. Wang, D. Xiao, and Q. Huang, “A novel multi-label pest image classifier using the modified swin transformer and soft binary cross entropy loss,” Engineering Applications of Artificial Intelligence , vol. 126, p. 107060, 2023

    work page 2023

  7. [6]

    A two- stream network with complementary feature fusion for pest image classification,

    C. Wang, J. Zhang, J. He, W. Luo, X. Yuan, and L. Gu, “A two- stream network with complementary feature fusion for pest image classification,” Engineering Applications of Artificial Intelligence , vol. 124, p. 106563, 2023

    work page 2023

  8. [7]

    An improved densenet model to classify the damage caused by cotton aphid,

    W. Bao, T. Cheng, X.-G. Zhou, W. Guo, Y . Wang, X. Zhang, H. Qiao, and D. Zhang, “An improved densenet model to classify the damage caused by cotton aphid,” Computers and Electronics in Agriculture, vol. 203, p. 107485, 2022

    work page 2022

Show all 35 references
  1. [8]

    Deep transfer learning model for disease identification in wheat crop,

    S. Nigam, R. Jain, S. Marwaha, A. Arora, M. A. Haque, A. Dheeraj, and V . K. Singh, “Deep transfer learning model for disease identification in wheat crop,” Ecological Informatics, vol. 75, p. 102068, 2023

    2023

  2. [9]

    Inception convolutional vision trans- formers for plant disease identification,

    S. Yu, L. Xie, and Q. Huang, “Inception convolutional vision trans- formers for plant disease identification,” Internet of Things , vol. 21, p. 100650, 2023

    2023

  3. [10]

    Ai based rice leaf disease identification enhanced by dynamic mode decomposition,

    S. KM, S. V , S. P. Kurian, and O. K. Sikha, “Ai based rice leaf disease identification enhanced by dynamic mode decomposition,” 2023

    2023

  4. [11]

    Trends in vision- based machine learning techniques for plant disease identification: A systematic review,

    P. S. Thakur, P. Khanna, T. Sheorey, and A. Ojha, “Trends in vision- based machine learning techniques for plant disease identification: A systematic review,” Expert Systems with Applications , p. 118117, 2022

    2022

  5. [12]

    A survey on smart agriculture: Development modes, tech- nologies, and security and privacy challenges,

    X. Yang, L. Shu, J. Chen, M. A. Ferrag, J. Wu, E. Nurellari, and K. Huang, “A survey on smart agriculture: Development modes, tech- nologies, and security and privacy challenges,” IEEE/CAA Journal of Automatica Sinica, vol. 8, no. 2, pp. 273–302, 2021

    2021

  6. [13]

    Rice-fusion: A multimodality data fusion framework for rice disease diagnosis,

    R. R. Patil and S. Kumar, “Rice-fusion: A multimodality data fusion framework for rice disease diagnosis,” IEEE Access, vol. 10, pp. 5207– 5222, 2022

    2022

  7. [14]

    Multimodal fine-grained transformer model for pest recognition,

    Y . Zhang, L. Chen, and Y . Yuan, “Multimodal fine-grained transformer model for pest recognition,” Electronics, vol. 12, no. 12, p. 2620, 2023

    2023

  8. [15]

    Crop disease identification and interpretation method based on multimodal deep learning,

    J. Zhou, J. Li, C. Wang, H. Wu, C. Zhao, and G. Teng, “Crop disease identification and interpretation method based on multimodal deep learning,” Computers and Electronics in Agriculture , vol. 189, p. 106408, 2021

    2021

  9. [16]

    Tomato disease classification and identification method based on multimodal fusion deep learning,

    N. Zhang, H. Wu, H. Zhu, Y . Deng, and X. Han, “Tomato disease classification and identification method based on multimodal fusion deep learning,” Agriculture, vol. 12, no. 12, p. 2014, 2022

    2014

  10. [17]

    Real-esrgan: Training real- world blind super-resolution with pure synthetic data,

    X. Wang, L. Xie, C. Dong, and Y . Shan, “Real-esrgan: Training real- world blind super-resolution with pure synthetic data,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2021, pp. 1905–1914

    2021

  11. [18]

    Super-resolution via deep learning,

    K. Hayat, “Super-resolution via deep learning,” arXiv preprint arXiv:1706.09077, 2017

    2017 arXiv

  12. [19]

    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

    2017

  13. [20]

    A guide to convolution arithmetic for deep learning,

    V . Dumoulin and F. Visin, “A guide to convolution arithmetic for deep learning,” arXiv preprint arXiv:1603.07285 , 2016

    2016 arXiv

  14. [21]

    Yolov4: Op- timal speed and accuracy of object detection,

    A. Bochkovskiy, C.-Y . Wang, and H.-Y . M. Liao, “Yolov4: Op- timal speed and accuracy of object detection,” arXiv preprint arXiv:2004.10934, 2020

    2004 arXiv

  15. [22]

    Spatial pyramid pooling in deep convolutional networks for visual recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Spatial pyramid pooling in deep convolutional networks for visual recognition,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 37, no. 9, pp. 1904– 1916, 2015

    1904

  16. [23]

    Path aggregation network for instance segmentation,

    S. Liu, L. Qi, H. Qin, J. Shi, and J. Jia, “Path aggregation network for instance segmentation,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , 2018, pp. 8759–8768

    2018

  17. [24]

    Yolov3: An incremental improvement,

    J. Redmon and A. Farhadi, “Yolov3: An incremental improvement,” arXiv preprint arXiv:1804.02767 , 2018

    2018 arXiv

  18. [25]

    Ip102: A large- scale benchmark dataset for insect pest recognition,

    X. Wu, C. Zhan, Y .-K. Lai, M.-M. Cheng, and J. Yang, “Ip102: A large- scale benchmark dataset for insect pest recognition,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019, pp. 8787–8796

    2019

  19. [26]

    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,” Advances in Neural Information Processing Systems , vol. 28, 2015

    2015

  20. [27]

    Feature pyramid networks for object detection,

    T.-Y . Lin, P. Doll´ar, R. Girshick, K. He, B. Hariharan, and S. Belongie, “Feature pyramid networks for object detection,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , 2017, pp. 2117–2125

    2017

  21. [28]

    Ssd: Single shot multibox detector,

    W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. Reed, C.-Y . Fu, and A. C. Berg, “Ssd: Single shot multibox detector,” in Computer Vision– ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part I 14 . Springer, 2016, pp. 21–37

    2016

  22. [30]

    Real-time flying object detection with yolov8,

    D. Reis, J. Kupec, J. Hong, and A. Daoudi, “Real-time flying object detection with yolov8,” 2024. [Online]. Available: https: //arxiv.org/abs/2305.09972

    2024 arXiv

  23. [31]

    Yolov9: Learning what you want to learn using programmable gradient information,

    C.-Y . Wang, I.-H. Yeh, and H.-Y . M. Liao, “Yolov9: Learning what you want to learn using programmable gradient information,” 2024. [Online]. Available: https://arxiv.org/abs/2402.13616

    2024 arXiv

  24. [32]

    Efficient estimation of word representations in vector space,

    T. Mikolov, K. Chen, G. Corrado, and J. Dean, “Efficient estimation of word representations in vector space,” 2013. [Online]. Available: https://arxiv.org/abs/1301.3781

    2013 arXiv

  25. [33]

    Enhanced litehrnet based sheep weight estimation using rgb-d images,

    C. He, Y . Qiao, R. Mao, M. Li, and M. Wang, “Enhanced litehrnet based sheep weight estimation using rgb-d images,”Computers and Electronics in Agriculture, vol. 206, p. 107667, 2023

    2023

  26. [34]

    Ex- plainable deep convolutional neural networks for insect pest recogni- tion,

    S. Coulibaly, B. Kamsu-Foguem, D. Kamissoko, and D. Traore, “Ex- plainable deep convolutional neural networks for insect pest recogni- tion,” Journal of Cleaner Production , vol. 371, p. 133638, 2022

    2022

  27. [35]

    Application of transfer learning and image augmentation technology for tomato pest identifi- cation,

    M.-L. Huang, T.-C. Chuang, and Y .-C. Liao, “Application of transfer learning and image augmentation technology for tomato pest identifi- cation,” Sustainable Computing: Informatics and Systems , vol. 33, p. 100646, 2022

    2022

Pith tools

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