Articles Open Access Published 26 June 2026

WaveGate-DETR: Wavelet-Gated Feature Fusion and Density-Aware Sparse Queries for Real-Time Tiny Pest Detection in UAV Imagery of Tropical Rice Fields

Authors

Rizky Ananda Putra1*, Dewi Kartika Sari2, Yoga Pramudya3, Ahmad Fauzan Hakim1

  1. Department of Informatics, Universitas Nusantara Digital, Salatiga, Indonesia
  2. Faculty of Agricultural Technology, Institut Teknologi Archipelago, Makassar, Indonesia
  3. Department of Electrical Engineering, Universitas Samudra Raya, Surabaya, Indonesia

* Corresponding author: rizky.putra@und.example

PDF Cite References

Abstract

Early detection of rice pests from low-altitude unmanned aerial vehicle (UAV) imagery can drastically reduce pesticide use in smallholder farms, yet the targets—golden apple snail egg masses, rice bugs and early stem-borer symptoms—typically occupy fewer than 16 × 16 pixels and appear in highly uneven clusters. Real-time DETR-style detectors lose these objects for two reasons: strided down-sampling discards the high-frequency cues that define tiny objects, and a fixed set of object queries is wasted on empty paddy while being insufficient for dense clusters. We propose WaveGate-DETR, a real-time end-to-end detector with two novel components. First, a Wavelet-Gated Feature Pyramid (WG-FPN) replaces lossy strided convolutions with a discrete wavelet transform and learns a spatial gate that re-injects high-frequency sub-bands only where tiny objects are likely, preserving edges without amplifying water-surface glare. Second, Density-Aware Sparse Queries (DASQ) predict an object-density map and allocate a variable number of queries per image, concentrating decoding capacity on clusters. We also introduce NusaPest-UAV, a new dataset of 4,812 high-resolution images with 61,350 annotated instances of six pest and damage classes collected in 12 regencies across Java and Sulawesi, Indonesia. WaveGate-DETR-S achieves 71.8% mAP@0.5 and 29.8% AP on tiny objects, improving RT-DETR-R18 by +5.7 and +7.9 points respectively, while using 68% fewer queries on average and running at 58 FPS on an NVIDIA Jetson Orin NX. On VisDrone2019-DET the method also generalises, reaching 50.3% mAP@0.5. Code, trained models and the dataset will be released to support reproducible research on agricultural vision in the tropics.

Keywords

How to Cite

R. A. Putra, D. K. Sari, Y. Pramudya, and A. F. Hakim, “WaveGate-DETR: Wavelet-Gated Feature Fusion and Density-Aware Sparse Queries for Real-Time Tiny Pest Detection in UAV Imagery of Tropical Rice Fields,” FEATURE: Nusantara Journal of Computer Vision, vol. 1, no. 1, pp. 1–7, Jun. 2026.

Author Biographies
  • Rizky Ananda Putra

    Lecturer and researcher at the Computer Vision Laboratory, Department of Informatics, Universitas Nusantara Digital. His interests include aerial object detection and efficient transformers.

  • Dewi Kartika Sari

    Researcher in precision agriculture and remote sensing at Institut Teknologi Archipelago.

  • Yoga Pramudya

    Assistant professor working on embedded deep learning and UAV systems.

  • Ahmad Fauzan Hakim

    Graduate researcher focusing on wavelet-based representation learning.

References (25)
  1. N. Carion, F. Massa, G. Synnaeve, N. Usunier, A. Kirillov, and S. Zagoruyko, "End-to-end object detection with transformers," in Proc. Eur. Conf. Comput. Vis. (ECCV), 2020, pp. 213–229.
  2. X. Zhu, W. Su, L. Lu, B. Li, X. Wang, and J. Dai, "Deformable DETR: Deformable transformers for end-to-end object detection," in Proc. Int. Conf. Learn. Represent. (ICLR), 2021.
  3. Y. Zhao, W. Lv, S. Xu, J. Wei, G. Wang, Q. Dang, Y. Liu, and J. Chen, "DETRs beat YOLOs on real-time object detection," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2024, pp. 16965–16974.
  4. H. Zhang, F. Li, S. Liu, L. Zhang, H. Su, J. Zhu, L. M. Ni, and H.-Y. Shum, "DINO: DETR with improved denoising anchor boxes for end-to-end object detection," in Proc. Int. Conf. Learn. Represent. (ICLR), 2023.
  5. G. Jocher, A. Chaurasia, and J. Qiu, "Ultralytics YOLOv8," 2023. [Online]. Available: https://github.com/ultralytics/ultralytics
  6. T.-Y. Lin, P. Dollár, R. Girshick, K. He, B. Hariharan, and S. Belongie, "Feature pyramid networks for object detection," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2017, pp. 2117–2125.
  7. P. Zhu, L. Wen, D. Du, X. Bian, H. Fan, Q. Hu, and H. Ling, "Detection and tracking meet drones challenge," IEEE Trans. Pattern Anal. Mach. Intell., vol. 44, no. 11, pp. 7380–7399, 2022.
  8. J. Wang, W. Yang, H. Guo, R. Zhang, and G.-S. Xia, "Tiny object detection in aerial images," in Proc. Int. Conf. Pattern Recognit. (ICPR), 2021, pp. 3791–3798.
  9. P. Liu, H. Zhang, K. Zhang, L. Lin, and W. Zuo, "Multi-level wavelet-CNN for image restoration," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. Workshops (CVPRW), 2018, pp. 773–782.
  10. Q. Li, L. Shen, S. Guo, and Z. Lai, "Wavelet integrated CNNs for noise-robust image classification," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2020, pp. 7245–7254.
  11. S. G. Mallat, "A theory for multiresolution signal decomposition: The wavelet representation," IEEE Trans. Pattern Anal. Mach. Intell., vol. 11, no. 7, pp. 674–693, 1989.
  12. P. Sun, R. Zhang, Y. Jiang, T. Kong, C. Xu, W. Zhan, M. Tomizuka, L. Li, Z. Yuan, C. Wang, and P. Luo, "Sparse R-CNN: End-to-end object detection with learnable proposals," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2021, pp. 14454–14463.
  13. F. Yang, H. Fan, P. Chu, E. Blasch, and H. Ling, "Clustered object detection in aerial images," in Proc. IEEE/CVF Int. Conf. Comput. Vis. (ICCV), 2019, pp. 8311–8320.
  14. C. Li, T. Yang, S. Zhu, C. Chen, and S. Guan, "Density map guided object detection in aerial images," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. Workshops (CVPRW), 2020, pp. 190–191.
  15. X. Wu, C. Zhan, Y.-K. Lai, M.-M. Cheng, and J. Yang, "IP102: A large-scale benchmark dataset for insect pest recognition," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2019, pp. 8787–8796.
  16. K. He, X. Zhang, S. Ren, and J. Sun, "Deep residual learning for image recognition," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2016, pp. 770–778.
  17. T.-Y. Lin, P. Goyal, R. Girshick, K. He, and P. Dollár, "Focal loss for dense object detection," in Proc. IEEE Int. Conf. Comput. Vis. (ICCV), 2017, pp. 2980–2988.
  18. H. Zhang, Y. Wang, F. Dayoub, and N. Sünderhauf, "VarifocalNet: An IoU-aware dense object detector," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2021, pp. 8514–8523.
  19. H. Rezatofighi, N. Tsoi, J. Gwak, A. Sadeghian, I. Reid, and S. Savarese, "Generalized intersection over union: A metric and a loss for bounding box regression," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2019, pp. 658–666.
  20. J. Wang, C. Xu, W. Yang, and L. Yu, "A normalized Gaussian Wasserstein distance for tiny object detection," arXiv:2110.13389, 2021.
  21. F. Li, H. Zhang, S. Liu, J. Guo, L. M. Ni, and L. Zhang, "DN-DETR: Accelerate DETR training by introducing query denoising," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2022, pp. 13619–13627.
  22. S. Liu, F. Li, H. Zhang, X. Yang, X. Qi, H. Su, J. Zhu, and L. Zhang, "DAB-DETR: Dynamic anchor boxes are better queries for DETR," in Proc. Int. Conf. Learn. Represent. (ICLR), 2022.
  23. D. P. Hughes and M. Salathé, "An open access repository of images on plant health to enable the development of mobile disease diagnostics," arXiv:1511.08060, 2015.
  24. A. Dosovitskiy et al., "An image is worth 16x16 words: Transformers for image recognition at scale," in Proc. Int. Conf. Learn. Represent. (ICLR), 2021.
  25. Z. Liu, Y. Lin, Y. Cao, H. Hu, Y. Wei, Z. Zhang, S. Lin, and B. Guo, "Swin transformer: Hierarchical vision transformer using shifted windows," in Proc. IEEE/CVF Int. Conf. Comput. Vis. (ICCV), 2021, pp. 10012–10022.