Enhancing Leaf Area Segmentation by Using Attention Gates and Knowledge Distillation in UNet Architecture
DOI:
https://doi.org/10.26636/jtit.2025.3.2079Keywords:
attention gate, knowledge distillation, modified light weight UNet, semantic segmentationAbstract
Accurate segmentation of leaf regions plays a vital role in plant phenotyping and agricultural analysis. This paper presents AKDUNet, a lightweight UNet-based architecture that integrates attention gates and knowledge distillation to improve segmentation performance while minimizing computational complexity. The architecture replaces traditional skip connections with attention gates to focus on salient spatial features and employs a two-stage training pipeline, where a compact student model learns from a deeper teacher model using a tailored distillation loss function. AKDUNet is evaluated on two benchmark datasets (CWFID and Sunflower) and outperforms a range of state-of-the-art models, including UNet++, Inception UNet, VGG-based UNets, SDUNet, INSCA UNet, and SegFormer. Ablation studies confirm the advantages of attention modules, and qualitative analyses using Grad-CAM visualizations reveal the model's ability to effectively focus on crucial leaf structures. The results demonstrate that AKDUNet is not only computationally efficient but also highly accurate, making it suitable for real-time deployment in resource-constrained agricultural environments.
Downloads
References
[1] A. Walter and U. Schurr, "The Modular Character of Growth in Nicotiana Tabacum Plants under Steady-state Nutrition", Journal of Experimental Botany, vol. 50, pp. 1169-1177, 1999.
View in Google Scholar
DOI: https://doi.org/10.1093/jxb/50.336.1169
[2] A.L. Chandra, S.V. Desai, W. Guo, and V.N. Balasubramanian, "Computer Vision with Deep Learning for Plant Phenotyping in Agriculture: A Survey", ArXiv, 2020 (https://doi.org/10.48550/arXiv.2006.11391).
View in Google Scholar
[3] S. Haug and J. Ostermann, "A Crop/weed Field Image Dataset for the Evaluation of Computer Vision Based Precision Agriculture Tasks", Proc. of Computer Vision - ECCV 2014, pp. 105-116, 2014.
View in Google Scholar
DOI: https://doi.org/10.1007/978-3-319-16220-1_8
[4] R. Liu et al., "TransKD: Transformer Knowledge Distillation for Efficient Semantic Segmentation", IEEE Transactions on Intelligent Transportation Systems, vol. 25, 2024.
View in Google Scholar
DOI: https://doi.org/10.1109/TITS.2024.3455416
[5] J. Zhang, Q. Liang, and Y. Shi, "KD-SCFNet: Towards More Accurate and Efficient Salient Object Detection via Knowledge Distillation", ArXiv, 2022.
View in Google Scholar
DOI: https://doi.org/10.2139/ssrn.4081836
[6] S. An, Q. Liao, Z. Lu, and J.-H. Xue, "Efficient Semantic Segmentation via Self-attention and Self-distillation", IEEE Transactions on Intelligent Transportation Systems, vol. 23, pp. 15256-15266, 2022.
View in Google Scholar
DOI: https://doi.org/10.1109/TITS.2021.3139001
[7] T. Zhang et al., "Efficient RGB-T Tracking via Cross-modality Distillation", IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, Canada, 2023.
View in Google Scholar
DOI: https://doi.org/10.1109/CVPR52729.2023.00523
[8] X. Hu, K. Yang, L. Fei, and K. Wang, "ACNet: Attention Based Network to Exploit Complementary Features for RGBD Semantic Segmentation", IEEE International Conference on Image Processing (ICIP), Taipei, Taiwan, 2019.
View in Google Scholar
DOI: https://doi.org/10.1109/ICIP.2019.8803025
[9] W. Zhou, J. Yuan, J. Lei, and T. Luo, "TSNet: Three-stream Self-attention Network for RGB-D Indoor Semantic Segmentation", IEEE Intelligent Systems, vol. 36, pp. 73-78, 2021.
View in Google Scholar
DOI: https://doi.org/10.1109/MIS.2020.2999462
[10] W. Zhou, E. Yang, J. Lei, and L. Yu, "FRNet: Feature Reconstruction Network for RGB-D Indoor Scene Parsing", IEEE Journal of Selected Topics in Signal Processing, vol. 16, pp. 677-687, 2022.
View in Google Scholar
DOI: https://doi.org/10.1109/JSTSP.2022.3174338
[11] J. Zhou et al., "ESA-Net: A Network with Efficient Spatial Attention for Smoky Vehicle Detection", 2021 IEEE International Instrumentation and Measurement Technology Conference (I2MTC), Glasgow, UK, 2021.
View in Google Scholar
DOI: https://doi.org/10.1109/I2MTC50364.2021.9460078
[12] P. Chavan, P.P. Chavan, and A. Chavan, "Hybrid Architecture for Crop Detection and Leaf Disease Detection with Improved U-Net Segmentation Model and Image Processing", Crop Protection, vol. 190, art. no. 107117, 2025.
View in Google Scholar
DOI: https://doi.org/10.1016/j.cropro.2025.107117
[13] M.K. Surehli, N. Aggarwal, G. Joshi, and H. Nayyar, "Semantic Segmentation of Plant Structures with Deep Learning and Channel-wise Attention Mechanism", JTIT, vol. 99, pp. 56-66, 2025.
View in Google Scholar
DOI: https://doi.org/10.26636/jtit.2025.1.1853
[14] A. Das et al., "Deep Learning-based Classification, Detection, and Segmentation of Tomato Leaf Diseases: A State-of-the-art Review", Artificial Intelligence in Agriculture, vol. 15, pp. 192-220, 2025 (https://doi.org/10.1016/j.aiia.2025.02.006).
View in Google Scholar
DOI: https://doi.org/10.1016/j.aiia.2025.02.006
[15] C. Guo et al., "SD-UNet: A Structured Dropout U-Net for Retinal Vessel Segmentation", 2019 IEEE 19th International Conference on Bioinformatics and Bioengineering (BIBE), Athens, Greece, 2019.
View in Google Scholar
DOI: https://doi.org/10.1109/BIBE.2019.00085
[16] I. Delibasoglu, "INCSA-UNET: Spatial Attention Inception UNET for Aerial Images Segmentation", Computing and Informatics, vol. 40, pp. 1244-1262, 2022.
View in Google Scholar
DOI: https://doi.org/10.31577/cai_2021_6_1244
[17] M. Agarwal, S.K. Gupta, and K.K. Biswas, "Plant Leaf Disease Segmentation Using Compressed UNet Architecture", Proc. of Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD), pp. 9-14, 2021.
View in Google Scholar
DOI: https://doi.org/10.1007/978-3-030-75015-2_2
[18] E. Xie et al., "SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers", ArXiv, 2021.
View in Google Scholar
[19] O. Ronneberger, P. Fischer, and T. Brox, "U-Net: Convolutional Networks for Biomedical Image Segmentation", Medical Image Computing and Computer-Assisted Intervention - MICCAI 2015, Munich, Germany, 2015.
View in Google Scholar
DOI: https://doi.org/10.1007/978-3-319-24574-4_28
[20] M. Fawakherji et al., "Multi-spectral Image Synthesis for Crop/weed Segmentation in Precision Farming", Robotics and Autonomous Systems, vol. 146, art. no. 103861, 2021.
View in Google Scholar
DOI: https://doi.org/10.1016/j.robot.2021.103861
[21] Z. Zhou, M.R. Siddiquee, N. Tajbakhsh, and J. Liang, "UNet++: A Nested UNet Architecture for Medical Image Segmentation", in Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support, Lecture Notes in Computer Science, Springer Cham, pp. 3-11, 2018.
View in Google Scholar
DOI: https://doi.org/10.1007/978-3-030-00889-5_1
[22] N.S. Punn, S. Agarwal, "Inception U-Net Architecture for Semantic Segmentation to Identify Nuclei in Microscopy Cell Images", ACM Transactions on Multimedia Computing, Communications, and Applications, vol. 16, no. 1, pp. 1-15, 2020.
View in Google Scholar
DOI: https://doi.org/10.1145/3376922
[23] Y. Miao et al., "CT Image Segmentation of Foxtail Millet Seeds Based on Semantic Segmentation Model VGG16-UNet", Plant Methods, vol. 20, art. no. 169, 2024.
View in Google Scholar
DOI: https://doi.org/10.1186/s13007-024-01288-y
[24] K. Simonyan and A. Zisserman, "Very Deep Convolutional Networks for Large-scale Image Recognition", ArXiv, 2014.
View in Google Scholar
[25] K. He, X. Zhang, S. Ren, and J. Sun, "Deep Residual Learning for Image Recognition", 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, USA, 2016.
View in Google Scholar
DOI: https://doi.org/10.1109/CVPR.2016.90
Downloads
Submitted
Accepted
Published
Issue
Section
License
Copyright (c) 2025 A. Shamim Banu, S. Deivalakshmi
This work is licensed under a Creative Commons Attribution 4.0 International License.
