scholarly journals Identification of Plant-Leaf Diseases Using CNN and Transfer-Learning Approach

Electronics ◽  
2021 ◽  
Vol 10 (12) ◽  
pp. 1388
Author(s):  
Sk Mahmudul Hassan ◽  
Arnab Kumar Maji ◽  
Michał Jasiński ◽  
Zbigniew Leonowicz ◽  
Elżbieta Jasińska
Keyword(s):  
Disease Classification ◽  
Batch Size ◽  
Healthy Plant ◽  
Plant Leaves ◽  
Early Prevention ◽  
Computation Cost ◽  
Training Time ◽  
Feature Based ◽  
Deep Cnn ◽  
Accuracy Rates

The timely identification and early prevention of crop diseases are essential for improving production. In this paper, deep convolutional-neural-network (CNN) models are implemented to identify and diagnose diseases in plants from their leaves, since CNNs have achieved impressive results in the field of machine vision. Standard CNN models require a large number of parameters and higher computation cost. In this paper, we replaced standard convolution with depth=separable convolution, which reduces the parameter number and computation cost. The implemented models were trained with an open dataset consisting of 14 different plant species, and 38 different categorical disease classes and healthy plant leaves. To evaluate the performance of the models, different parameters such as batch size, dropout, and different numbers of epochs were incorporated. The implemented models achieved a disease-classification accuracy rates of 98.42%, 99.11%, 97.02%, and 99.56% using InceptionV3, InceptionResNetV2, MobileNetV2, and EfficientNetB0, respectively, which were greater than that of traditional handcrafted-feature-based approaches. In comparison with other deep-learning models, the implemented model achieved better performance in terms of accuracy and it required less training time. Moreover, the MobileNetV2 architecture is compatible with mobile devices using the optimized parameter. The accuracy results in the identification of diseases showed that the deep CNN model is promising and can greatly impact the efficient identification of the diseases, and may have potential in the detection of diseases in real-time agricultural systems.

scholarly journals Training Tips for the Transformer Model

2018 ◽  
Vol 110 (1) ◽  
pp. 43-70 ◽  
Author(s):  
Martin Popel ◽  
Ondřej Bojar
Keyword(s):  
Sentence Length ◽  
Batch Size ◽  
Critical Parameters ◽  
Neural Machine Translation ◽  
Training Time ◽  
Multiple Gpus ◽  
Translation Quality ◽  
Transformer Model ◽  
Data Constraints ◽  
And Training

Abstract This article describes our experiments in neural machine translation using the recent Tensor2Tensor framework and the Transformer sequence-to-sequence model (Vaswani et al., 2017). We examine some of the critical parameters that affect the final translation quality, memory usage, training stability and training time, concluding each experiment with a set of recommendations for fellow researchers. In addition to confirming the general mantra “more data and larger models”, we address scaling to multiple GPUs and provide practical tips for improved training regarding batch size, learning rate, warmup steps, maximum sentence length and checkpoint averaging. We hope that our observations will allow others to get better results given their particular hardware and data constraints.

scholarly journals Disease classification in Solanum melongena using deep learning

2019 ◽  
Vol 17 (3) ◽  
pp. e0204 ◽  
Author(s):  
Krishnaswamy R. Aravind ◽  
Purushothaman Raja ◽  
Rajendran Ashiwin ◽  
Konnaiyar V. Mukesh
Keyword(s):  
Deep Learning ◽  
Tamil Nadu ◽  
Solanum Melongena ◽  
Disease Classification ◽  
Healthy Plant ◽  
Future Research ◽  
Learning Models ◽  
Cercospora Leaf Spot ◽  
Before And After ◽  
Augmentation Techniques

Aim of study: The application of pre-trained deep learning models, AlexNet and VGG16, for classification of five diseases (Epilachna beetle infestation, little leaf, Cercospora leaf spot, two-spotted spider mite and Tobacco Mosaic Virus (TMV)) and a healthy plant in Solanum melongena (brinjal in Asia, eggplant in USA and aubergine in UK) with images acquired from smartphones.Area of study: Images were acquired from fields located at Alangudi (Pudukkottai district), Tirumalaisamudram and Pillayarpatti (Thanjavur district) – Tamil Nadu, India.Material and methods: Most of earlier studies have been carried out with images of isolated leaf samples, whereas in this work the whole or part of the plant images were utilized for the dataset creation. Augmentation techniques were applied to the manually segmented images for increasing the dataset size. The classification capability of deep learning models was analysed before and after augmentation. A fully connected layer was added to the architecture and evaluated for its performance.Main results: The modified architecture of VGG16 trained with the augmented dataset resulted in an average validation accuracy of 96.7%. Despite the best accuracy, all the models were tested with sample images from the field and the modified VGG16 resulted in an accuracy of 93.33%.Research highlights: The findings provide a guidance for possible factors to be considered in future research relevant to the dataset creation and methodology for efficient prediction using deep learning models.

scholarly journals Optimizing Deep Convolutional Neural Network for Facial Expression Recognition

2020 ◽  
Vol 5 (2) ◽  
pp. 192-195
Author(s):  
Umesh B. Chavan ◽  
Dinesh Kulkarni
Keyword(s):  
Neural Network ◽  
Facial Expression ◽  
Convolutional Neural Network ◽  
Facial Expression Recognition ◽  
Deep Convolutional Neural Network ◽  
Expression Recognition ◽  
Graphic Processing Units ◽  
Data Set ◽  
Training Time ◽  
Deep Cnn

Facial expression recognition (FER) systems have attracted much research interest in the area of Machine Learning. We designed a large, deep convolutional neural network to classify 40,000 images in the data-set into one of seven categories (disgust, fear, happy, angry, sad, neutral, surprise). In this project, we have designed deep learning Convolution Neural Network (CNN) for facial expression recognition and developed model in Theano and Caffe for training process. The proposed architecture achieves 61% accuracy. This work presents results of accelerated implementation of the CNN with graphic processing units (GPUs). Optimizing Deep CNN is to reduce training time for system.

scholarly journals Face Verification with Multi-Task and Multi-Scale Features Fusion

2017 ◽  
Author(s):  
Xiaojun Lu ◽  
Yue Yang ◽  
Weilin Zhang ◽  
Qi Wang ◽  
Yang Wang
Keyword(s):  
Training Dataset ◽  
Face Verification ◽  
Deep Convolutional Neural Networks ◽  
Computation Cost ◽  
Face Representation ◽  
Features Fusion ◽  
Multi Scale ◽  
The Face ◽  
In The Wild ◽  
Deep Cnn

Face verification for unrestricted faces in the wild is a challenging task. This paper proposes a method based on two deep convolutional neural networks(CNN) for face verification. In this work, we explore to use identification signal to supervise one CNN and the combination of semi-verification and identification to train the other one. In order to estimate semi-verification loss at a low computation cost, a circle, which is composed of all faces, is used for selecting face pairs from pairwise samples. In the process of face normalization, we propose to use different landmarks of faces to solve the problems caused by poses. And the final face representation is formed by the concatenating feature of each deep CNN after PCA reduction. What's more, each feature is a combination of multi-scale representations through making use of auxiliary classifiers. For the final verification, we only adopt the face representation of one region and one resolution of a face jointing Joint Bayesian classifier. Experiments show that our method can extract effective face representation with a small training dataset and our algorithm achieves 99.71% verification accuracy on LFW dataset.

scholarly journals Performance Measures of Respiratory Disease Classification using Deep Learning

2021 ◽  
Vol 9 (VI) ◽  
pp. 884-892
Author(s):  
G. Rama Janani
Keyword(s):  
Neural Network ◽  
Feature Extraction ◽  
Deep Learning ◽  
Performance Measures ◽  
Respiratory Infections ◽  
Respiratory Illness ◽  
Extraction Methods ◽  
Disease Classification ◽  
Deep Cnn

The paper is based on classification of respiratory illness like covid 19 and pneumonia by using deep learning. The symptoms of COVID-19 and pneumonia are similar. Due to this, it is often difficult to identify what is causing your condition without being tested for COVID-19 or other respiratory infections. To find out how COVID-19 and pneumonia differs from one another, this paper presents that a novel Convolutional Neural Network in Tensor Flow and Keras based Covid-19 pneumonia classification. The proposed system supported implements CNN using Pneumonia images to classify the Covid-19, normal, pneumonia. The knowledge from these studies can potentially help in diagnosis of the concerned disease. It is predicted that the success of the anticipated results will increase if the CNN method is supported by adding extra feature extraction methods for classifying covid-19 and pneumonia successfully thereby improving the efficacy and potential of using deep CNN to pictures.

scholarly journals Classification of Apple Disease Based on Non-Linear Deep Features

2021 ◽  
Vol 11 (14) ◽  
pp. 6422
Author(s):  
Hamail Ayaz ◽  
Erick Rodríguez-Esparza ◽  
Muhammad Ahmad ◽  
Diego Oliva ◽  
Marco Pérez-Cisneros ◽  
...  
Keyword(s):  
High Volume ◽  
Disease Classification ◽  
Baseline Model ◽  
Agricultural Industry ◽  
Machine Learning Methods ◽  
Proposed Model ◽  
Conventional Machine ◽  
Deep Cnn ◽  
Comparative Results

Diseases in apple orchards (rot, scab, and blotch) worldwide cause a substantial loss in the agricultural industry. Traditional hand picking methods are subjective to human efforts. Conventional machine learning methods for apple disease classification depend on hand-crafted features that are not robust and are complex. Advanced artificial methods such as Convolutional Neural Networks (CNN’s) have become a promising way for achieving higher accuracy although they need a high volume of samples. This work investigates different Deep CNN (DCNN) applications to apple disease classification using deep generative images to obtain higher accuracy. In order to achieve this, our work progressively modifies a baseline model by using an end-to-end trained DCNN model that has fewer parameters, better recognition accuracy than existing models (i.e., ResNet, SqeezeNet, and MiniVGGNet). We have performed a comparative study with state-of-the-art CNN as well as conventional methods proposed in the literature, and comparative results confirm the superiority of our proposed model.

scholarly journals A Chaotic Salp Swarm Feature Selection Algorithm for Apple and Tomato Plant Leaf Disease Detection

2021 ◽  
Vol 10 (5) ◽  
pp. 3037-3045
Keyword(s):  
Deep Learning ◽  
Short Term Memory ◽  
Detection System ◽  
Plant Diseases ◽  
Healthy Plant ◽  
Disease Detection ◽  
Test Accuracy ◽  
Plant Leaves ◽  
Leaf Disease ◽  
Deep Learning Network

Plants are seen as vital because they provide mankind with energy. Plant diseases can harm the leaf at any time between planting and harvesting, resulting in enormous losses in crop output and market value. A leaf disease detection system acts asignificant role in agricultural production. A large amount of labour is required for this process as well as an in-depth understanding of plant diseases. Determining the presence of illnesses in plant leaves requires the use of deep learning and machine learning methods, which classify the data based on a specified set. In this paper, apple and tomato leaves disease detection process is carried out by Chaotic Salp Swarm algorithm (CSSA) followed by Bi-directional Long Short Term Memory (Bi-LSTM) technique for classification. We've used the Bi-LSTM architecture to sense disease in tomato and apple leaves in studies. In order to determine the type of leaves, we trained a deep learning network using the PlantVillage dataset of damaged and healthy plant leaves. It is estimated that the trained model achieves a test accuracy of 96%.

scholarly journals Flawless Identification of Fusarium Oxysporum in Tomato Plant Leaves by Machine Learning Algorithm

2021 ◽  
Vol 2 (4) ◽  
pp. 194-201
Author(s):  
Dhaya R
Keyword(s):  
Fusarium Oxysporum ◽  
Tomato Plant ◽  
Learning Algorithm ◽  
Visual Observation ◽  
Fungal Diseases ◽  
Plant Diseases ◽  
Cultivated Plants ◽  
Healthy Plant ◽  
Plant Leaves ◽  
Sweet Potatoes

In the olden days, plant diseases could be measured by visual observation and based on the level and severity of the symptoms on plant leaves. Over the day, it became a high-level degree of complexity due to the huge volume of cultivated plants. Now a day, the diseases are very different due to diverted manure procedures, and its diagnosis will be very tough even experienced farmers and agronomists too. Even though, after diagnosis, there is a lack of perfect remedy or mistaken treatment for that. The plants are affecting by many vascular fungal diseases which are widespread in many crops. Fusarium wilt (FW) is one of the fungal diseases in many plants. Mostly the tomato, sweet potatoes, tobacco, legumes, cucurbits plants are affected by this Fusarium oxysporum (FO) disease often due to its soil. The main goal of this research article is used to determine FO disease in the tomato plant leaves. Besides, the proposed algorithm constructs model with two times classifying and identifying the disease for better accuracy. The open database consists of 87k images with 60% affected leaves images, 40% healthy plant leaves too. Our proposed hybrid algorithm is found the disease with 96% accuracy with the huge amount of dataset.

scholarly journals Training-Time-Friendly Network for Real-Time Object Detection

2020 ◽  
Vol 34 (07) ◽  
pp. 11685-11692
Author(s):  
Zili Liu ◽  
Tu Zheng ◽  
Guodong Xu ◽  
Zheng Yang ◽  
Haifeng Liu ◽  
...  
Keyword(s):  
Real Time ◽  
State Of The Art ◽  
Batch Size ◽  
Training Process ◽  
Training Time ◽  
Sample Weights ◽  
Gaussian Kernels ◽  
Novel Approach ◽  
Training Samples ◽  
Speed And Accuracy

Modern object detectors can rarely achieve short training time, fast inference speed, and high accuracy at the same time. To strike a balance among them, we propose the Training-Time-Friendly Network (TTFNet). In this work, we start with light-head, single-stage, and anchor-free designs, which enable fast inference speed. Then, we focus on shortening training time. We notice that encoding more training samples from annotated boxes plays a similar role as increasing batch size, which helps enlarge the learning rate and accelerate the training process. To this end, we introduce a novel approach using Gaussian kernels to encode training samples. Besides, we design the initiative sample weights for better information utilization. Experiments on MS COCO show that our TTFNet has great advantages in balancing training time, inference speed, and accuracy. It has reduced training time by more than seven times compared to previous real-time detectors while maintaining state-of-the-art performances. In addition, our super-fast version of TTFNet-18 and TTFNet-53 can outperform SSD300 and YOLOv3 by less than one-tenth of their training time, respectively. The code has been made available at https://github.com/ZJULearning/ttfnet.

Sign in / Sign up

Export Citation Format

Share Document