DLMC-Net: Deeper lightweight multi-class classification model for plant leaf disease detection

Sharma, Vivek; Tripathi, Ashish; Mittal, Himanshu

doi:10.1016/j.ecoinf.2023.102025

Cited by 56 publications

(14 citation statements)

References 73 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Nevertheless, the training sets have exhibited some disparities between the projected and real labels due to unforeseen circumstances resulting from errors during annotation. Still, through rigorous experimentation, we achieved an impressive accuracy of around 98% and minimal validation loss compared with other CNN models (57)(58)(59)(60)(61). As a part of model validation, we further deployed ResNet50v2 as a part of the application for utilization by grape farmers in the study area.…”

Section: Discussionmentioning

confidence: 88%

Grape (Vitis Vinifera) Leaf Disease Detection and Classification Using Deep Learning Techniques: A Study on Real-Time Grape Leaf Image Dataset in India

Shah,

Kumbhar,

Singh

2024

IJE

View full text Add to dashboard Cite

In modern horticulture, the grape industry across the globe has been coping with the issue of grape crop diseases. The detection of grape leaf diseases using automated methods can greatly assist farmers in mitigating yield losses and ensuring sustainability. However, existing systems face hurdles while handling grape leaf images at the farm level, and these models fail to generalize well on un-seen images. This study proposes the development of a well-curated real-time dataset of grape leaf images assimilated through field visits in the study area in India. This designed dataset is further used to train convolutional neural network models to accurately identify and classify grape leaves as either diseased or healthy. The potential of transfer learning using CNN models like VGG, ResNet, Inception, and Xception is assessed on the curated dataset. Our findings indicate that ResNet50V2 outperformed the other models in accurately identifying and classifying grape leaf diseases. Using transfer learning, existing weights (pretrained) and learned features were utilized for further training and fine-tuning the CNN models on our curated dataset. The results of the proposed approach are compared with existing automated grape leaf disease identification techniques. It is observed that the proposed approach, which is on a real-time grape leaf image dataset, provides the highest accuracy among others. Further, this study provides a wellcurated dataset of on-field grape leaf images in the Indian context, which can serve as a benchmark for future research. This study shows that deep learning techniques can aid farmers in identifying grape leaf diseases early.

show abstract

Section: Discussionmentioning

confidence: 88%

Grape (Vitis Vinifera) Leaf Disease Detection and Classification Using Deep Learning Techniques: A Study on Real-Time Grape Leaf Image Dataset in India

Shah,

Kumbhar,

Singh

2024

IJE

View full text Add to dashboard Cite

show abstract

“…The proposed model demonstrates high accuracy, with a success rate of 99.26%. Sharma et al (2023) introduced DLMC-Net, a more profound lightweight convolutional neural network architecture, for real-time plant leaf disease detection across multiple crops. The suggested model extracts deep features using a succession of collective blocks and the passage layer.…”

Section: Related Workmentioning

confidence: 99%

A robust and light-weight transfer learning-based architecture for accurate detection of leaf diseases across multiple plants using less amount of images

Mazumder,

Mridha,

Alfarhood

et al. 2024

Front. Plant Sci.

View full text Add to dashboard Cite

Leaf diseases are a global threat to crop production and food preservation. Detecting these diseases is crucial for effective management. We introduce LeafDoc-Net, a robust, lightweight transfer-learning architecture for accurately detecting leaf diseases across multiple plant species, even with limited image data. Our approach concatenates two pre-trained image classification deep learning-based models, DenseNet121 and MobileNetV2. We enhance DenseNet121 with an attention-based transition mechanism and global average pooling layers, while MobileNetV2 benefits from adding an attention module and global average pooling layers. We deepen the architecture with extra-dense layers featuring swish activation and batch normalization layers, resulting in a more robust and accurate model for diagnosing leaf-related plant diseases. LeafDoc-Net is evaluated on two distinct datasets, focused on cassava and wheat leaf diseases, demonstrating superior performance compared to existing models in accuracy, precision, recall, and AUC metrics. To gain deeper insights into the model’s performance, we utilize Grad-CAM++.

show abstract

“…Their proposed network worked quite well on test dataset and obtained 99.1% classification accuracy. Sharma et al. (2023) designed a lightweight DLMC-Net model by using novel collective blocks and passage layers.…”

Section: Related Workmentioning

confidence: 99%

TrIncNet: a lightweight vision transformer network for identification of plant diseases

Gole¹,

Bedi²,

Marwaha³

et al. 2023

Front. Plant Sci.

View full text Add to dashboard Cite

In the agricultural sector, identifying plant diseases at their earliest possible stage of infestation still remains a huge challenge with respect to the maximization of crop production and farmers’ income. In recent years, advanced computer vision techniques like Vision Transformers (ViTs) are being successfully applied to identify plant diseases automatically. However, the MLP module in existing ViTs is computationally expensive as well as inefficient in extracting promising features from diseased images. Therefore, this study proposes a comparatively lightweight and improved vision transformer network, also known as “TrIncNet” for plant disease identification. In the proposed network, we introduced a modified encoder architecture a.k.a. Trans-Inception block in which the MLP block of existing ViT was replaced by a custom inception block. Additionally, each Trans-Inception block is surrounded by a skip connection, making it much more resistant to the vanishing gradient problem. The applicability of the proposed network for identifying plant diseases was assessed using two plant disease image datasets viz: PlantVillage dataset and Maize disease dataset (contains in-field images of Maize diseases). The comparative performance analysis on both datasets reported that the proposed TrIncNet network outperformed the state-of-the-art CNN architectures viz: VGG-19, GoogLeNet, ResNet-50, Xception, InceptionV3, and MobileNet. Moreover, the experimental results also showed that the proposed network had achieved 5.38% and 2.87% higher testing accuracy than the existing ViT network on both datasets, respectively. Therefore, the lightweight nature and improved prediction performance make the proposed network suitable for being integrated with IoT devices to assist the stakeholders in identifying plant diseases at the field level.

show abstract

DLMC-Net: Deeper lightweight multi-class classification model for plant leaf disease detection

Cited by 56 publications

References 73 publications

Grape (Vitis Vinifera) Leaf Disease Detection and Classification Using Deep Learning Techniques: A Study on Real-Time Grape Leaf Image Dataset in India

Grape (Vitis Vinifera) Leaf Disease Detection and Classification Using Deep Learning Techniques: A Study on Real-Time Grape Leaf Image Dataset in India

A robust and light-weight transfer learning-based architecture for accurate detection of leaf diseases across multiple plants using less amount of images

TrIncNet: a lightweight vision transformer network for identification of plant diseases

Contact Info

Product

Resources

About