Enhancing the performance of transferred efficientnet models in leaf image-based plant disease classification

Hanh, Bui Thi; Manh, Hoang Van; Nguyen, Ngoc-Viet

doi:10.1007/s41348-022-00601-y

Cited by 23 publications

(8 citation statements)

References 55 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(1) Transfer learning Transfer learning is a deep-learning method that aims to accelerate a model's training process (13). Therefore, for this study, we decided to use the ResNet152V2 pre-trained model from the 2015 ImageNet competition for transfer learning.…”

Section: Resnet152v2-sa-se Network Constructionmentioning

confidence: 99%

“…The EfficientNet-B0 model achieved scores for class accuracy of 98.33, 96.51, 95.45, 100, 100, 99.26, and 98.72% for healthy, full chalky, chalky discolored, half chalky, broken, discolored, and normal damage classes, respectively ( 12 ). At present, pre-training models based on transfer learning are increasingly being utilized by academics, and can minimize the training time for models ( 13 ). However, deep-learning models are still deficient, and it is difficult to control the attention of models in the task of image recognition.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Intelligent grading method for walnut kernels based on deep learning and physiological indicators

Chen¹,

Dai²,

Jian³

et al. 2023

Front. Nutr.

View full text Add to dashboard Cite

Walnut grading is an important step before the product enters the market. However, traditional walnut grading primarily relies on manual assessment of physiological features, which is difficult to implement efficiently. Furthermore, walnut kernel grading is, at present, relatively unsophisticated. Therefore, this study proposes a novel deep-learning model based on a spatial attention mechanism and SE-network structure to grade walnut kernels using machine vision to ensure accuracy and improve assessment efficiency. In this experiment, we found through the literature that both the lightness (L* value) and malondialdehyde (MDA) contens of walnut kernels were correlated with the oxidation phenomenon in walnuts. Subsequently, we clustered four partitionings using the L* values. We then used the MDA values to verify the rationality of these partitionings. Finally, four network models were used for comparison and training: VGG19, EfficientNetB7, ResNet152V2, and spatial attention and spatial enhancement network combined with ResNet152V2 (ResNet152V2-SA-SE). We found that the ResNet152V2-SA-SE model exhibited the best performance, with a maximum test set accuracy of 92.2%. The test set accuracy was improved by 6.2, 63.2, and 74.1% compared with that of ResNet152V2, EfficientNetB7, and VGG19, respectively. Our testing demonstrated that combining spatial attention and spatial enhancement methods improved the recognition of target locations and intrinsic information, while decreasing the attention given to non-target regions. Experiments have demonstrated that combining spatial attention mechanisms with SE networks increases focus on recognizing target locations and intrinsic information, while decreasing focus on non-target regions. Finally, by comparing different learning rates, regularization methods, and batch sizes of the model, we found that the training performance of the model was optimal with a learning rate of 0.001, a batch size of 128, and no regularization methods. In conclusion, this study demonstrated that the ResNet152V2-SA-SE network model was effective in the detection and evaluation of the walnut kernels.

show abstract

Section: Resnet152v2-sa-se Network Constructionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Intelligent grading method for walnut kernels based on deep learning and physiological indicators

Chen¹,

Dai²,

Jian³

et al. 2023

Front. Nutr.

View full text Add to dashboard Cite

show abstract

“…The first convolution operation, with a kernel size of 1*1, was utilized to increase the dimension of the input feature matrix. MBConv6 in Table 1 signified that the scale of convolution kernels was 6 times that of the The feature learning of the impurity-containing maize images was conducted through eight convolution stages; as shown in Table 1, the width and depth of each stage were closely related to the dimension of the original images, which were obtained by multiplying the magnification factor corresponding to the resolution with the parameters of the baseline (EfficientNetB0) [45,46] (where H i * W i * C i are the dimensions of the feature matrix before operation O i in Figure 2). L i denotes the quantity of repetitions of the operation O i , i.e., the depth of stage i.…”

Section: Image Feature Learning Networkmentioning

confidence: 99%

A Tiny Object Detection Approach for Maize Cleaning Operations

Yu,

Li,

et al. 2023

Foods

View full text Add to dashboard Cite

Real-time and accurate awareness of the grain situation proves beneficial for making targeted and dynamic adjustments to cleaning parameters and strategies, leading to efficient and effective removal of impurities with minimal losses. In this study, harvested maize was employed as the raw material, and a specialized object detection network focused on impurity-containing maize images was developed to determine the types and distribution of impurities during the cleaning operations. On the basis of the classic contribution Faster Region Convolutional Neural Network, EfficientNetB7 was introduced as the backbone of the feature learning network and a cross-stage feature integration mechanism was embedded to obtain the global features that contained multi-scale mappings. The spatial information and semantic descriptions of feature matrices from different hierarchies could be fused through continuous convolution and upsampling operations. At the same time, taking into account the geometric properties of the objects to be detected and combining the images’ resolution, the adaptive region proposal network (ARPN) was designed and utilized to generate candidate boxes with appropriate sizes for the detectors, which was beneficial to the capture and localization of tiny objects. The effectiveness of the proposed tiny object detection model and each improved component were validated through ablation experiments on the constructed RGB impurity-containing image datasets.

show abstract

“…EfficientNet (Tan & Le, 2019) is a novel convolutional neural network architecture for analyzing images and is preferred as a prediction model in various disease diagnostic research for humans (Venugopal, Joseph, Das, & Nath, 2022) (Nayak, Padhy, Mallick, Zymbler, & Kumar, 2022) (Wang, Liu, Xie, Yang, & Zhou, 2021) (Ravi, Acharya, & Alazab, 2022) (Zhu et al, 2022) (Marques, Ferreras, & de la Torre-Diez, 2022) and plants (Hanh, Van Manh, & Nguyen, 2022) (Farman et al, 2022) (Atila, Uçar, Akyol, & Uçar, 2021) (Li, Liu, Li, & Liu, 2022). EfficientNet's popularity among deep learning researchers is due to its high classification and prediction performances.…”

Section: Literature Reviewmentioning

confidence: 99%

Diagnosing Diseases From Fingernail Images

CAN

Işık

2022

Eskişehir Osmangazi Üniversitesi Mühendislik Ve Mimarlık Fakültesi Dergisi

View full text Add to dashboard Cite

This paper investigates how people's finger and nail appearance helps diagnose various diseases, such as Darier's disease, Muehrcke's lines, alopecia areata, beau's lines, bluish nails, and clubbing, by image processing and deep learning techniques. We used a public dataset consisting of 17 different classes with 655 samples. We divided the dataset into three folds based on a widely used rule, the 0.7:0.2:0.1, for training, validation, and testing purposes. We tested the EfficientNet-B2 model for performance evaluation purposes by using Noisy-Student weights by setting the batch size and epochs as 32 and 1000. The model achieves a 72% accuracy score and 91% AUC score for test samples to detect fingernail diseases. The empirical findings in this study provide a new understanding that the EfficientNet-B2 model can categorize nail disease types through numerous classes.

show abstract

Enhancing the performance of transferred efficientnet models in leaf image-based plant disease classification

Cited by 23 publications

References 55 publications

Intelligent grading method for walnut kernels based on deep learning and physiological indicators

Intelligent grading method for walnut kernels based on deep learning and physiological indicators

A Tiny Object Detection Approach for Maize Cleaning Operations

Diagnosing Diseases From Fingernail Images

Contact Info

Product

Resources

About