植物病蟲害是農業領域的一大問題。如果不能及時解決或者判斷錯誤，容易導致作物品質不佳，產量下降，或者因為錯誤使用農藥導致環境污染。本研究將植物病蟲害辨識技術與科技結合，旨在通過科技手段節省人力資源，提高作物品質。在植物病蟲害辨識方面的研究中，大多數研究都是通過分析葉面上的病徵，使用CNN模型進行植物病蟲害辨識。然而，在實際的農田中，植物通常會感染多種疾病，因此葉面上會出現多種疾病病徵。如果只使用一般的CNN模型進行辨識，只能識別出一種病蟲害，這在實際情況下是不足夠的。
為了解決這個問題，我們提出了一個可以處理多葉片的Multi-label植物病蟲害辨識系統。該系統分為兩部分，第一部分為一個葉片檢測系統，名為Leaf Detector，Leaf Detector以YOLOX作為主架構，負責檢測圖像中的多片葉子；第二部分為一個multi-label的辨識系統，並名為Efficient Attention Multi-Task Network(EAMTNet)，EAMTNet先透過network backbone提取整張葉片特徵，再利用Multi-task learning概念，將特徵共享到Multi-task learning Branch上，透過Multi-task learning Branch對於某個疾病再去做特徵提取，最終生成該葉片的最終多標籤辨識結果。
在本文的實驗中，我們比較了本文提出的系統與其他multi-label或是single-label classification model的表現。實驗結果證明，相對於其他方法，我們提出的方法具有更好的準確率，在multi-label classification實驗中，我們的準確度可以比其他multi-label classification model高出約2%；而single-label classification實驗中，準確度大多也高於其他single-label classification model 1%-2%。同時，在相近的準確率下，本文提出的架構參數低於其他辨識系統模型約1/3。
本系統可大大提高植物病蟲害識別的效率和準確性，這對於確保作物產量和糧食安全至關重要。由於能夠處理多張葉子圖像並準確識別多種病蟲害，該系統可能成為農民和研究人員等方面的寶貴工具。此外，使用具有更少參數的更高效的體系結構可以使系統更易於訪問且具有成本效益以供廣泛使用。該系統的進一步研究和開發可能會導致植物病理學領域的重大進步。

In agriculture, plant pests and diseases are major problems. Failure to address these issues promptly or making incorrect judgments can lead to poor crop quality or environmental pollution caused by incorrect chemical application. While most CNN models for plant pest and disease identification analyze disease on the leaf surface to determine whether the plant is infected with a certain disease or not, after visiting actual farmlands, we discovered that plants are often not infected with just one disease, and multiple disease symptoms can be present on the leaf surface. Using only a general CNN model to identify pests or diseases is inadequate for this situation.
To deal with this problem, we propose a Multi-label Plant Disease and Pest Identification System that can handle multiple leaf images. The system consists of two parts: the first part is a leaf detection system called Leaf Detector, which uses YOLOX as the main framework to detect multiple leaf images in the input; the second part is a multi-label recognition system called Efficient Attention Multi-Task Network (EAMTNet). EAMTNet first extracts the overall leaf features through a network backbone and then utilizes the concept of Multi-task learning to share the features to the Multi-task learning Branch. The Multi-task learning Branch performs feature extraction for a specific disease, resulting in the final multi-label recognition result for that leaf image.
In our experiments, we compared the performance of our proposed system with other multi-label or single-label classification models. The results demonstrate that our method achieves better accuracy compared to other methods. In multi-label classification experiments, our accuracy is approximately 2% higher than other multi-label classification models, while in single-label classification experiments, the accuracy is mostly 1%-2% higher than other single-label classification models. Furthermore, our proposed architecture has approximately 1/3 fewer parameters compared to other recognition system models while achieving similar accuracy.
This proposed system has the potential to greatly improve the efficiency and accuracy of plant disease and pest identification, which is critical for ensuring crop yield and food security. With the ability to handle multiple leaf images and accurately identify multiple diseases and pests, this system could be a valuable tool for farmers and researchers alike. Additionally, the use of a more efficient architecture with fewer parameters could make the system more accessible and cost-effective for widespread use. Further research and development of this system could lead to significant advancements in the field of plant pathology.

摘要 I
ABSTRACT III
TABLE OF CONTENTS V
LIST OF TABLES VII
LIST OF FIGURE VIII
LIST OF FOMULA IX
1. INTRODUCTION 1
2. RELATED WORK 3
2.1. PLANT PEST AND DISEASE IDENTIFICATION 3
2.2. OBJECT DETECTION 5
2.3. IMAGE CLASSIFICATION 6
2.4. LIGHTWEIGHT NEURAL NETWORKS 7
2.5. ATTENTION MECHANISM 8
2.6. MULTI-LABEL LEARNING 10
2.7. MULTI-TASK LEARNING 10
3. PROPOSED METHOD 13
3.1. LEAF DETECTOR 14
3.2. EFFICIENT ATTENTION MULTI-TASK NETWORK 17
3.2.1 Multi-Task Learning 17
3.2.2 Network Backbone 18
3.2.3 Multi-Task Learning Branch 23
3.3 MTL BRANCH DESIGN 24
4. EXPERIMENTAL RESULTS 27
4.1. EXPERIMENTAL ENVIRONMENT 27
4.2. LOSS FUNCTION 27
4.3. EVALUATE METRIC 28
4.4. DATASET 29
4.4.1 Pomelo 29
4.4.2 PlantVillage 33
4.5 EXPERIMENTS ON POMELO-ML 35
4.5.1 Experiments on Different Optimizers 37
4.5.2 Experiments on Different Attention Models 39
4.5.3 Experiments on Different Learning Rates 41
4.6 EXPERIMENTS ON POMELO-SL 44
4.7 EXPERIMENTS ON PLANTVILLAGE 45
4.8 PERFORMANCE EVALUATE 47
5 CONCLUSION 49
REFERENCES 51

[1]E. Suryawati, R. Sustika, R. S. Yuwana, A. Subekti, and H. F. Pardede, "Deep Structured Convolutional Neural Network for Tomato Diseases Detection," in 2018 International Conference on Advanced Computer Science and Information Systems, pp. 385-390, IEEE, 2018.
[2]S. Sladojevic, M. Arsenovic, A. Anderla, D. Culibrk, and D. Stefanovic, "Deep Neural Networks Based Recognition of Plant Diseases by Leaf Image Classification," Computational Intelligence and Neuroscience, vol. 2016, p. 3289801, 2016.
[3]M. R. Howlader, U. Habiba, R. H. Faisal, and M. M. Rahman, "Automatic Recognition of Guava Leaf Diseases Using Deep Convolution Neural Network," in 2019 International Conference on Electrical, Computer and Communication Engineering, pp. 1-5, IEEE, 2019.
[4]S. P. Mohanty, D. P. Hughes, and M. Salathé, "Using Deep Learning for Image-Based Plant Disease Detection," Frontiers in Plant Science, vol. 7, no. 1419, Sep. 22, 2016.
[5]G. Wang, Y. Sun, and J. Wang, "Automatic Image-Based Plant Disease Severity Estimation Using Deep Learning," Computational Intelligence and Neuroscience, vol. 2017, 2017.
[6]B. A. Ashqar and S. S. Abu-Naser, "Image-Based Tomato Leaves Diseases Detection Using Deep Learning," International Journal of Academic Engineering Research, vol. 2, no. 12, pp. 10-16, 2018.
[7]E. C. Too, L. Yujian, S. Njuki, and L. Yingchun, "A Comparative Study of Fine-Tuning Deep Learning Models for Plant Disease Identification," Computers and Electronics in Agriculture, vol. 161, pp. 272-279, 2019.
[8]M. H. Saleem, J. Potgieter, and K. M. Arif, "Plant Disease Classification: A Comparative Evaluation of Convolutional Neural Networks and Deep Learning Optimizers," Plants, vol. 9, no. 10, p. 1319, Oct. 2020.
[9]L. Tan, J. Lu, and H. Jiang, "Tomato Leaf Diseases Classification Based on Leaf Images: A Comparison between Classical Machine Learning and Deep Learning Methods," AgriEngineering, vol. 3, no. 3, pp. 542-558, Jul. 2021.
[10]Y. Borhani, J. Khoramdel, and E. Najafi, "A deep learning based approach for automated plant disease classification using vision transformer," Sci Rep, vol. 12, p. 11554, 2022.
[11]R. Girshick, J. Donahue, T. Darrell, J. Malik, "Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation," in 2014 IEEE Conference on Computer Vision and Pattern Recognition, 2014, pp. 580-587.
[12]J. Uijlings, K. van de Sande, T. Gevers, and A. Smeulders, "Selective Search for Object Recognition," IJCV2013, 2013, pp. 154-171.
[13]R. Girshick, "Fast R-CNN," in 2015 IEEE International Conference on Computer Vision (ICCV), 2015, pp. 1440-1448.
[14]K. He, X. Zhang, S. Ren, J. Sun, "Spatial pyramid pooling in deep convolutional networks for visual recognition," in IEEE transactions on pattern analysis and machine intelligence, vol. 37, no. 9, 2015, pp. 1904-1916.
[15]S. Ren, K. He, R. Girshick, J. Sun, "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks," in IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 39, no. 6, 2017, pp. 1137-1149.
[16]T.-Y. Lin, P. Dollár, R. Girshick, K. He, B. Hariharan, S. Belongie, "Feature pyramid networks for object detection," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2017..
[17]Z. Cai, N. Vasconcelos, "Cascade R-CNN: Delving into High Quality Object Detection," in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2018.
[18]J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, "You Only Look Once: Unified, Real-Time Object Detection," in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016.
[19]W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. Reed, C.-Y. Fu, A. C. Berg, "SSD: Single Shot Multibox Detector," in European Conference on Computer Vision, Springer, Cham, 2016.
[20]J. Redmon, A. Farhadi, "YOLO9000: better, faster, stronger." J. Redmon, A. Farhadi, "YOLO9000: better, faster, stronger," in Proceedings of the IEEE conference on computer vision and pattern recognition., 2017.
[21]T.-Y. Lin, P. Goyal, R. Girshick, K. He, P. Dollár, "Focal loss for dense object detection," in Proceedings of the IEEE International Conference on Computer Vision, 2017.
[22]J. Redmon, A. Farhadi, "YOLOv3: An incremental improvement," arXiv preprint arXiv:1804.02767, 2018.
[23]M. Tan, R. Pang, Q. V. Le, "EfficientDet: Scalable and efficient object detection," in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020.
[24]A. Bochkovskiy, C. Y. Wang, and H. Y. M. Liao, "Yolov4: Optimal speed and accuracy of object detection," arXiv preprint arXiv:2004.10934, 2020.
[25]Z. Ge, S. Liu, F. Wang, Z. Li, and J. Sun, "Yolox: Exceeding YOLO Series in 2021," arXiv preprint arXiv:2107.08430, 2021.
[26]Y. Lecun, L. Bottou, Y. Bengio and P. Haffner, "Gradient-based learning applied to document recognition," in Proceedings of the IEEE, vol. 86, no. 11, 1998, pp. 2278-2324, doi: 10.1109/5.726791.
[27]A. Krizhevsky, I. Sutskever, and G. E. Hinton, "Imagenet Classification with Deep Convolutional Neural Networks," in Advances in neural information processing systems, 2012, pp. 1097-1105.
[28]H. Guo, S. Wang, "Long-Tailed Multi-Label Visual Recognition by Collaborative Training on Uniform and Re-Balanced Samplings.", in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021, pp. 15089-15098
[29]K. Simonyan, A. Zisserman, "Very Deep Convolutional Networks for Large-Scale Image Recognition," in The International Conference on Learning Representations, 2015.
[30]C. Szegedy, W. Liu, Y. Jia, P. Sermanet, S. Reed, D. Anguelov, D. Erhan, V. Vanhoucke, A. Rabinovich, "Going Deeper with Convolutions," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2015, pp. 1-9.
[31]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), 2016, pp. 770-778, doi: 10.1109/CVPR.2016.90.
[32]G. Huang, Z. Liu, L. Van Der Maaten, K. Q. Weinberger, "Densely Connected Convolutional Networks," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp. 4700-4708.
[33]Y. Huang, Y. Cheng, D. Chen, H. Lee, J. Ngiam, Q.V. Le, Z. Chen "Gpipe: Efficient training of giant neural networks using pipeline parallelism," arXiv preprint arXiv:1808.07233, 2018.
[34]M. Tan, Q. Le, "EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks," in Proceedings of the 36th International Conference on Machine Learning, 2019, pp. 6105-6114.
[35]M. Tan, Q. Le, "EfficientNetV2: Smaller Models and Faster Training," in Proceedings of the 38th International Conference on Machine Learning, 2021, pp. 10096-10106.
[36]F. N. Iandola, S. Han, M.W. Moskewicz, K. Ashraf, W.J. Dally, and K. Keutzer, "SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and< 0.5 MB model size," arXiv preprint arXiv:1602.07360., 2016.
[37]A. G. Howard, M. Zhu, B. Chen, D. Kalenichenko, W. Wang, T. Weyand, M. Andreetto, H. Adam, "Mobilenets: Efficient convolutional neural networks for mobile vision applications," arXiv preprint arXiv:1704.04861, 2017.
[38]X. Zhang, X. Zhou, M. Lin, J. Sun, “ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2018.
[39]A. Gholami, K. Kwon, B. Wu, Z. Tai, X. Yue, P. Jin, S. Zhao, and K. Keutzer, “SqueezeNext: Hardware-Aware Neural Network Design,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops, 2018.
[40]N. Ma, X. Zhang, H.-T. Zheng, J. Sun, “ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design,” in Proceedings of the European conference on computer vision (ECCV), 2018.
[41]M. Sandler, A. Howard, M. Zhu, A. Zhmoginov, L. -C. Chen, "MobileNetV2: Inverted Residuals and Linear Bottlenecks," in 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018, pp. 4510-4520
[42]M. Tan, B. Chen, R. Pang, V. Vasudevan, M. Sandler, A. Howard, Q. V. Le, “MnasNet: Platform-Aware Neural Architecture Search for Mobile,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019.
[43]A. Howard, M. Sandler, G. Chu, L.-C. Chen, B. Chen, M. Tan, W. Wang, Y. Zhu, R. Pang, V. Vasudevan, Q. V. Le, H. Adam, “Searching for MobileNetV3,” in ICCV2019, 2019.
[44]Y. Li, Y. Chen, X. Dai, D. Chen, M. Liu, L. Yuan, Z. Liu, L. Zhang, N. Vasconcelos, “MicroNet: Towards Image Recognition with Extremely Low FLOPs,” arXiv preprint arXiv:2011.12289, 2020.
[45]F. Wang, M. Jiang, C. Qian, S. Yang, C. Li, H. Zhang, X. Wang, X. Tang, "Residual attention network for image classification," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2017.
[46]J. Hu, L. Shen, S. Albanie, G. Sun, E. Wu, "Squeeze-and-Excitation Networks," in 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018, pp. 7132-7141, doi: 10.1109/CVPR.2018.00745.
[47]X. Wang, R. Girshick, A. Gupta, K. He, "Non-local Neural Networks," in 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018, pp. 7794-7803, doi: 10.1109/CVPR.2018.00813.
[48]S. Woo, J. Park, J. Lee, I.S. Kweon, "CBAM: Convolutional block attention module," in Proceedings of the European conference on computer vision (ECCV), 2018.
[49]J. Fu, J. Liu, H. Tian, Y. Li, Y. Bao, Z. Fang, H. Lu, "Dual attention network for scene segmentation," in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019.
[50]Z. Huang, X. Wang, Y. Wei, L. Huang, H. Shi, W. Liu, T. S. Huang, "CCNet: Criss-cross attention for semantic segmentation," in Proceedings of the IEEE/CVF international conference on computer vision, 2019.
[51]Y. Cao, J. Xu, S. Lin, F. Wei, H. Hu, "GCnet: Non-local networks meet squeeze-excitation networks and beyond," in Proceedings of the IEEE/CVF international conference on computer vision workshops, 2019.
[52]Q. Wang, B. Wu, P. Zhu, P. Li, W. Zuo, Q. Hu, "ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks," CVPR2020, 2020.
[53]M. L. Zhang, Y. K. Li, X. Y. Liu, X. Geng, "Binary relevance for multi-label learning: an overview," Frontiers of Computer Science volume 12, 2018, pp. 191-202
[54]M. Zhang, Z. Zhou, "A Review on Multi-Label Learning Algorithms," in IEEE Transactions on Knowledge and Data Engineering, vol. 26, 2014, pp. 1819-1837
[55]A. N. Tarekegn, M. Giacobini, K. Michalak, "A review of methods for imbalanced multi-label classification," Pattern Recognition Volume 118, 2021.
[56]Y. Wei, W. Xia, J. Huang, B. Ni, J. Dong, Y. Zhao, S. Yan, "CNN: Single-label to multi-label," arXiv preprint arXiv:1406.5726, 2014.
[57]J. Zhang, Q. Wu, C. Shen, J. Zhang, J. Lu, "Multilabel image classification with regional latent semantic dependencies," in IEEE Transactions on Multimedia 20.10, 2018.
[58]S. M. García, C. J. Mantas, J. G. Castellano, J. Abellán, "Using Credal C4.5 for Calibrated Label Ranking in Multi-Label Classification," in International Journal of Approximate Reasoning Volume 147, 2022, pp. 60-77
[59]R. Caruana, "Multitask Learning," Machine learning, vol. 28, no. 1, pp. 41-75, 1997.
[60]Y. Sun, Y. Chen, X. Wang, and X. Tang, "Deep Learning Face Representation by Joint Identification-Verification," in Advances in neural information processing systems, pp. 1988-1996, 2014.
[61]A. Kendall, Y. Gal, and R. Cipolla, "Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics," in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 7482-7491, 2018.
[62]I. Kokkinos, "UberNet: Training a Universal Convolutional Neural Network for Low-, Mid-, and High-Level Vision Using Diverse Datasets and Limited Memory," CVPR 2017, 2017, pp. 5454-5463.
[63]I. Misra, A. Shrivastava, A. Gupta, M. Hebert, "Cross-stitch networks for multi-task learning," CVPR 2016 , 2016.
[64]Y. Gao, J. Ma, M. Zhao, W. Liu, A. L. Yuille, "NDDR-CNN: Layerwise Feature Fusing in Multi-Task CNNs by Neural Discriminative Dimensionality Reduction," CVPR 2019, 2019, pp. 3200-3209.
[65]S. Liu, E. Johns, and A. J. Davison, “End-to-end multi-task learning with attention,” CVPR 2019, 2019.
[66]D. Xu, W. Ouyang, X. Wang, N. Sebe, "PAD-Net: Multi-tasks guided prediction-and-distillation network for simultaneous depth estimation and scene parsing," CVPR 2018, 2018.
[67]H. Zhang, M. Cisse, Y. N. Dauphin, and D. Lopez-Paz, "mixup: Beyond empirical risk minimization," in ICLR, 2018.
[68]G. Song, Y. Liu, and X. Wang, "Revisiting the sibling head in object detector," in CVPR, 2020.
[69]Y. Wu, Y. Chen, L. Yuan, Z. Liu, L. Wang, H. Li, and Y. Fu, "Rethinking classification and localization for object detection," in CVPR, 2020.
[70]M. Ji, K. Zhang, Q. Wu, and Z. J. S. C. Deng, "Multi-Label Learning for Crop Leaf Diseases Recognition and Severity Estimation Based on Convolutional Neural Networks," Soft Computing, vol. 24, no. 20, pp. 15327-15340, 2020.
[71]黃宣哲, "A Plant Diseases and Pests Recognition System based on Multi-label Learning,", 2021