Getinet Yilma
ABSTRACT
Deep convolutional neural network (DCNN) image analysis improved plant health management. DCNN methods require train samples with diverse feature distribution and discriminative features for the models to generalize well on unseen categories. The limited train sample coined with the locality nature of the convolution operator and commonly shared kernel weights led CNN models to a sub-optimal solution. To address these challenges, we trained a multi-head self-attention augmented CNN (AAC) and explored other attention augmented networks to jointly exploit global salient and local shared features for fine-grained plant disease classification and visualization. We further proposed stochastic train sample transformations to enlarge and improve train sample diversity and feature distribution. The experiment evaluations on the PlantVillage dataset confirmed that the attention augmentation methods demonstrated robust performance in disease classification and visualization. Specifically, the AAC network scored 99.56% with stochastic transformed train samples, 98.01% with simple geometric transformed samples, and 97.34% without transformation. Therefore, attention augmentation in general, AAC in particular, and stochastic sample transformation can address discriminative plant disease classification and visualization.
- 1. Tagel Aboneh, Abebe Rorissa, Ramasamy Srinivasagan, and Ashenafi Gemechu. 2021. Computer Vision Framework for Wheat Disease Identification and Classification Using Jetson GPU Infrastructure. Technologies 9, 3: 47. https://doi.org/10.3390/technologies9030047Google Scholar
Cross Ref
- 2. Marko Arsenovic, Mirjana Karanovic, Srdjan Sladojevic, Andras Anderla, and Darko Stefanovic. 2019. Solving Current Limitations of Deep Learning Based Approaches for Plant Disease Detection. Symmetry 11, 7: 939. https://doi.org/10.3390/sym11070939Google ScholarCross Ref
- 3. Irwan Bello, Barret Zoph, Quoc Le, Ashish Vaswani, and Jonathon Shlens. 2019. Attention augmented convolutional networks. In Proceedings of the IEEE International Conference on Computer Vision, 3285–3294. https://doi.org/10.1109/ICCV.2019.00338Google ScholarCross Ref
- 4. Quan Huu Cap, Hiroyuki Uga, Satoshi Kagiwada, and Hitoshi Iyatomi. 2020. LeafGAN: An Effective Data Augmentation Method for Practical Plant Disease Diagnosis. IEEE Transactions on Automation Science and Engineering: 1–10.Google ScholarCross Ref
- 5. Junde Chen, Defu Zhang, Md Suzauddola, Yaser Ahangari Nanehkaran, and Yuandong Sun. 2021. Identification of plant disease images via a squeeze-and-excitation MobileNet model and twice transfer learning. IET Image Processing 15, 5: 1115–1127.Google ScholarCross Ref
- 6. Ting Chen, Simon Kornblith, Mohammad Norouzi, and Geoffrey Hinton. 2020. A simple framework for contrastive learning of visual representations. In 37th International Conference on Machine Learning, ICML 2020, 1575–1585.Google ScholarDigital Library
- 7. Halil Durmus, Ece Olcay Gunes, and Murvet Kirci. 2017. Disease detection on the leaves of the tomato plants by using deep learning. In 2017 6th International Conference on Agro-Geoinformatics, Agro-Geoinformatics 2017, 1–5. https://doi.org/10.1109/Agro-Geoinformatics.2017.8047016Google ScholarCross Ref
- 8. Kumie Gedamu, Yanli Ji, Yang Yang, Ling Ling Gao, and Heng Tao Shen. 2021. Arbitrary-view human action recognition via novel-view action generation. Pattern Recognition 118, 2021: 108043. https://doi.org/10.1016/j.patcog.2021.108043Google Scholar
- 9. Jie Hu, Li Shen, and Gang Sun. 2018. Squeeze-and-Excitation Networks. In Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 7132–7141. https://doi.org/10.1109/CVPR.2018.00745Google Scholar
- 10. Gilbert Gutabaga Hungilo, Gahizi Emmanuel, and Andi W.R. Emanuel. 2019. Image processing techniques for detecting and classification of plant disease – A review. ACM International Conference Proceeding Series: 48–52. https://doi.org/10.1145/3332340.3332341Google ScholarDigital Library
- 11. DIng Jiang, Fudong Li, Yuequan Yang, and Song Yu. 2020. A Tomato Leaf Diseases Classification Method Based on Deep Learning. In Proceedings of the 32nd Chinese Control and Decision Conference, CCDC 2020, 1446–1450. https://doi.org/10.1109/CCDC49329.2020.9164457Google ScholarCross Ref
- 12. R. Karthik, M. Hariharan, Sundar Anand, Priyanka Mathikshara, Annie Johnson, and R Menaka. 2020. Attention embedded residual CNN for disease detection in tomato leaves. Applied Soft Computing Journal 86. https://doi.org/10.1016/j.asoc.2019.105933Google ScholarDigital Library
- 13. Sharada P. Mohanty, David P. Hughes, and Marcel Salathé. 2016. Using Deep Learning for Image-Based Plant Disease Detection. Frontiers in Plant Science 7, September: 1–10. https://doi.org/10.3389/fpls.2016.01419Google Scholar
- 14. Aravind Krishnaswamy Rangarajan, Raja Purushothaman, and Aniirudh Ramesh. 2018. Tomato crop disease classification using pre-trained deep learning algorithm. In Procedia Computer Science, 1040–1047. https://doi.org/10.1016/j.procs.2018.07.070Google Scholar
- 15. Davinder Singh, Naman Jain, Pranjali Jain, Pratik Kayal, Sudhakar Kumawat, and Nipun Batra. 2020. PlantDoc: A dataset for visual plant disease detection. ACM International Conference Proceeding Series: 249–253. https://doi.org/10.1145/3371158.3371196Google ScholarDigital Library
- 16. Sanghyun Woo, Jongchan Park, Joon-Young Lee, and In So Kweon. 2018. CBAM: Convolutional Block Attention Module. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 3–19. https://doi.org/10.1007/978-3-030-01234-2_1Google ScholarDigital Library
- 17. Getinet Yilma, Seid Belay, Kumie Gedamu, Maregu Assefa, and Melese Ayalew. 2021. Attention augmented residual network for tomato disease detection and classification. TURKISH JOURNAL OF ELECTRICAL ENGINEERING & COMPUTER SCIENCES 29, 2021/SI–1: 2869–2885. https://doi.org/10.3906/elk-2105-115Google ScholarCross Ref
- 18. Getinet Yilma, Seid Belay, Zhiguang Qin, Kumie Gedamu, and Melese Ayalew. 2020. Plant Disease Classification Using Two Pathway Encoder GAN Data Generation. In 2020 17th International Computer Conference on Wavelet Active Media Technology and Information Processing (ICCWAMTIP), 67–72. https://doi.org/10.1109/ICCWAMTIP51612.2020.9317494Google Scholar
- 19. Getinet Yilma, Kumie Gedamu, Maregu Assefa, Ariyo Oluwasanmi, and Zhiguang Qin. 2021. Generation and Transformation Invariant Learning for Tomato Disease Classification. In 2021 IEEE 2nd International Conference on Pattern Recognition and Machine Learning (PRML), 121–128. https://doi.org/10.1109/prml52754.2021.9520693Google Scholar
- 20. Hengshuang Zhao, Jiaya Jia, and Vladlen Koltun. 2020. Exploring Self-attention for Image Recognition. In Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 10073–10082. https://doi.org/10.1109/CVPR42600.2020.01009Google ScholarCross Ref
Index Terms
- Attention Augmented Convolutional Neural Network for Fine-Grained Plant Disease Classification and Visualization Using Stochastic Sample Transformations
Recommendations
Leveraging Fine-Grained Labels to Regularize Fine-Grained Visual Classification
ICCMS '19: Proceedings of the 11th International Conference on Computer Modeling and SimulationFine-grained visual categorization (FGVC) is challenging mainly due to the large intra-class confusion and small inter-class variance in terms of shape, pose, and appearance. We propose the concept of fine-grained label and that any given label can be ...
Convolutional low-resolution fine-grained classification
Highlights- Investigate fine-grained image classification from low-resolution images.
- ...
AbstractSuccessful fine-grained image classification methods learn subtle details between visually similar (sub-)classes, but the problem becomes significantly more challenging if the details are missing due to low resolution. Encouraged by ...
SEViT: a large-scale and fine-grained plant disease classification model based on transformer and attention convolution
AbstractPlant diseases are the leading cause of crop yield reduction. Rapid diagnosis using deep learning-based methods can effectively control the deterioration and spread of diseases. Convolutional Neural Network (CNN)-based methods are the current ...
Comments