scholarly journals Small-Scale Depthwise Separable Convolutional Neural Networks for Bacteria Classification

Electronics ◽  
2021 ◽  
Vol 10 (23) ◽  
pp. 3005
Author(s):  
Duc-Tho Mai ◽  
Koichiro Ishibashi
Keyword(s):  
Neural Networks ◽  
Convolutional Neural Networks ◽  
Recognition Accuracy ◽  
Computational Cost ◽  
Limited Resource ◽  
Vital Role ◽  
Small Scale ◽  
Automated Classification ◽  
Improving Accuracy ◽  
Low Computational Cost

Bacterial recognition and classification play a vital role in diagnosing disease by determining the presence of large bacteria in the specimens and the symptoms. Artificial intelligence and computer vision widely applied in the medical domain enable improving accuracy and reducing the bacterial recognition and classification time, which aids in making clinical decisions and choosing the proper treatment. This paper aims to provide an approach of 33 bacteria strains’ automated classification from the Digital Images of Bacteria Species (DIBaS) dataset based on small-scale depthwise separable convolutional neural networks. Our five-layer architecture has significant advantages due to the compact model, low computational cost, and reliable recognition accuracy. The experimental results proved that the proposed design reached the highest accuracy of 96.28% with a total of 6600 images and can be executed on limited-resource devices of 3.23 million parameters and 40.02 million multiply–accumulate operations (MACs). The number of parameters in this architecture is seven times less than the smallest model listed in the literature.

scholarly journals Improved Handwritten Digit Recognition Using Convolutional Neural Networks (CNN)

Sensors ◽  
2020 ◽  
Vol 20 (12) ◽  
pp. 3344 ◽  
Author(s):  
Savita Ahlawat ◽  
Amit Choudhary ◽  
Anand Nayyar ◽  
Saurabh Singh ◽  
Byungun Yoon
Keyword(s):  
Neural Networks ◽  
Convolutional Neural Networks ◽  
Character Recognition ◽  
Optical Character Recognition ◽  
Recognition Accuracy ◽  
Handwriting Recognition ◽  
Computational Cost ◽  
Handwritten Digit Recognition ◽  
Digit Recognition ◽  
Handwritten Digit

Traditional systems of handwriting recognition have relied on handcrafted features and a large amount of prior knowledge. Training an Optical character recognition (OCR) system based on these prerequisites is a challenging task. Research in the handwriting recognition field is focused around deep learning techniques and has achieved breakthrough performance in the last few years. Still, the rapid growth in the amount of handwritten data and the availability of massive processing power demands improvement in recognition accuracy and deserves further investigation. Convolutional neural networks (CNNs) are very effective in perceiving the structure of handwritten characters/words in ways that help in automatic extraction of distinct features and make CNN the most suitable approach for solving handwriting recognition problems. Our aim in the proposed work is to explore the various design options like number of layers, stride size, receptive field, kernel size, padding and dilution for CNN-based handwritten digit recognition. In addition, we aim to evaluate various SGD optimization algorithms in improving the performance of handwritten digit recognition. A network’s recognition accuracy increases by incorporating ensemble architecture. Here, our objective is to achieve comparable accuracy by using a pure CNN architecture without ensemble architecture, as ensemble architectures introduce increased computational cost and high testing complexity. Thus, a CNN architecture is proposed in order to achieve accuracy even better than that of ensemble architectures, along with reduced operational complexity and cost. Moreover, we also present an appropriate combination of learning parameters in designing a CNN that leads us to reach a new absolute record in classifying MNIST handwritten digits. We carried out extensive experiments and achieved a recognition accuracy of 99.87% for a MNIST dataset.

scholarly journals Optimizing 3D Convolution Kernels on Stereo Matching for Resource Efficient Computations

Sensors ◽  
2021 ◽  
Vol 21 (20) ◽  
pp. 6808
Author(s):  
Jianqiang Xiao ◽  
Dianbo Ma ◽  
Satoshi Yamane
Keyword(s):  
Neural Networks ◽  
Computational Complexity ◽  
Convolutional Neural Networks ◽  
Stereo Matching ◽  
State Of The Art ◽  
Computational Cost ◽  
The State ◽  
Matching Network ◽  
Convolution Kernels ◽  
Low Computational Cost

Despite recent stereo matching algorithms achieving significant results on public benchmarks, the problem of requiring heavy computation remains unsolved. Most works focus on designing an architecture to reduce the computational complexity, while we take aim at optimizing 3D convolution kernels on the Pyramid Stereo Matching Network (PSMNet) for solving the problem. In this paper, we design a series of comparative experiments exploring the performance of well-known convolution kernels on PSMNet. Our model saves the computational complexity from 256.66G MAdd (Multiply-Add operations) to 69.03G MAdd (198.47G MAdd to 10.84G MAdd for only considering 3D convolutional neural networks) without losing accuracy. On Scene Flow and KITTI 2015 datasets, our model achieves results comparable to the state-of-the-art with a low computational cost.

Web application to support evidence of individual emotional impact evoked by COVID-19 pandemic restrictions (Preprint)

2021 ◽  
Author(s):  
Hugo Mitre-Hernandez ◽  
Rodolfo Ferro-Perez ◽  
Francisco Gonzalez-Hernandez
Keyword(s):  
Neural Network ◽  
Mental Health ◽  
Neural Networks ◽  
Emotion Recognition ◽  
Web Application ◽  
Deep Neural Network ◽  
Data Transfer ◽  
Computational Cost ◽  
Low Computational Cost ◽  
The Web

BACKGROUND Mental health effects during COVID-19 quarantine need to be handled because patients, relatives, and healthcare workers are living with negative emotional behaviors. The clinical disorders of depression and anxiety are evoking anger, fear, sadness, disgust, and reducing happiness. Therefore, track emotions with the help of psychologists on online consultations –to reduce the risk of contagion– will go a long way in assisting with mental health. The human micro-expressions can describe genuine emotions of people and can be captured by Deep Neural Networks (DNNs) models. But the challenge is to implement it under the poor performance of a part of society's computers and the low speed of internet connection. OBJECTIVE This study aimed to create a useful and usable web application to record emotions in a patient’s card in real-time, achieving a small data transfer, and a Convolutional Neural Networks (CNN) model with a low computational cost. METHODS To validate the low computational cost premise, firstly, we compare DNN architectures results, collecting the floating-point operations per second (FLOPS), the Number of Parameters (NP) and accuracy from the MobileNet, PeleeNet, Extended Deep Neural Network (EDNN), Inception- Based Deep Neural Network (IDNN) and our proposed Residual mobile-based Network (ResmoNet) model. Secondly, we compare the trained models' results in terms of Main Memory Utilization (MMU) and Response Time to complete the Emotion recognition (RTE). Finally, we design a data transfer that includes the raw data of emotions and the basic text information of the patient. The web application was evaluated with the System Usability Scale (SUS) and a utility questionnaire by psychologists and psychiatrists (experts). RESULTS All CNN models were set up using 150 epochs for training and testing comparing the results for each variable in ResmoNet with the best model. It was obtained that ResmoNet has 115,976 NP less than MobileNet, 243,901 FLOPS less than MobileNet, and 5% less accuracy than EDNN (95%). Moreover, ResmoNet used less MMU than any model, only EDNN overcomes ResmoNet in 0.01 seconds for RTE. Finally, with our model, we develop a web application to collect emotions in real-time during a psychological consultation. For data transfer, the patient’s card and raw emotional data have 2 kb with a UTF-8 encoding approximately. Finally, according to the experts, the web application has good usability (73.8 of 100) and utility (3.94 of 5). CONCLUSIONS A usable and useful web application for psychologists and psychiatrists is presented. This tool includes an efficient and light facial emotion recognition model. Its purpose is to be a complementary tool for diagnostic processes.

Convolutional Neural Networks for the Localization of Plastic Velocity Gradient Tensor in Polycrystalline Microstructures

2021 ◽  
pp. 1-41
Author(s):  
David Montes de Oca Zapiain ◽  
Apaar Shanker ◽  
Surya Kalidindi
Keyword(s):  
Neural Networks ◽  
Convolutional Neural Networks ◽  
Velocity Gradient ◽  
Optimization Problems ◽  
Computational Cost ◽  
Two Phase ◽  
Design And Optimization ◽  
Gradient Fields ◽  
Generalized Spherical Harmonics ◽  
New Research

Abstract Recent work has demonstrated the potential of convolutional neural networks (CNNs) in producing low-computational cost surrogate models for the localization of mechanical fields in two-phase microstructures. The extension of the same CNNs to polycrystalline microstructures is hindered by the lack of an efficient formalism for the representation of the crystal lattice orientation in the input channels of the CNNs. In this paper, we demonstrate the benefits of using generalized spherical harmonics (GSH) for addressing this challenge. A CNN model was successfully trained to predict the local plastic velocity gradient fields in polycrystalline microstructures subjected to a macroscopically imposed loading condition. Specifically, it is demonstrated that the proposed approach improves significantly the accuracy of the CNN models, when compared with the direct use of Bunge-Euler angles to represent the crystal orientations in the input channels. Since the proposed approach implicitly satisfies the expected crystal symmetries in the specification of the input microstructure to the CNN, it opens new research directions for the adoption of CNNs in addressing a broad range of polycrystalline microstructure design and optimization problems.

scholarly journals An Efficient Algorithm for Cardiac Arrhythmia Classification Using Ensemble of Depthwise Separable Convolutional Neural Networks

2020 ◽  
Vol 10 (2) ◽  
pp. 483 ◽  
Author(s):  
Eko Ihsanto ◽  
Kalamullah Ramli ◽  
Dodi Sudiana ◽  
Teddy Surya Gunawan
Keyword(s):  
Neural Networks ◽  
Feature Extraction ◽  
Cardiac Arrhythmia ◽  
Convolutional Neural Networks ◽  
Computational Cost ◽  
Training Data ◽  
Qrs Detection ◽  
Convolutional Network ◽  
Novel Method ◽  
Electrocardiogram Ecg

Many algorithms have been developed for automated electrocardiogram (ECG) classification. Due to the non-stationary nature of the ECG signal, it is rather challenging to use traditional handcraft methods, such as time-based analysis of feature extraction and classification, to pave the way for machine learning implementation. This paper proposed a novel method, i.e., the ensemble of depthwise separable convolutional (DSC) neural networks for the classification of cardiac arrhythmia ECG beats. Using our proposed method, the four stages of ECG classification, i.e., QRS detection, preprocessing, feature extraction, and classification, were reduced to two steps only, i.e., QRS detection and classification. No preprocessing method was required while feature extraction was combined with classification. Moreover, to reduce the computational cost while maintaining its accuracy, several techniques were implemented, including All Convolutional Network (ACN), Batch Normalization (BN), and ensemble convolutional neural networks. The performance of the proposed ensemble CNNs were evaluated using the MIT-BIH arrythmia database. In the training phase, around 22% of the 110,057 beats data extracted from 48 records were utilized. Using only these 22% labeled training data, our proposed algorithm was able to classify the remaining 78% of the database into 16 classes. Furthermore, the sensitivity ( S n ), specificity ( S p ), and positive predictivity ( P p ), and accuracy ( A c c ) are 99.03%, 99.94%, 99.03%, and 99.88%, respectively. The proposed algorithm required around 180 μs, which is suitable for real time application. These results showed that our proposed method outperformed other state of the art methods.

scholarly journals Dealing with Lack of Training Data for Convolutional Neural Networks: The Case of Digital Pathology

Electronics ◽  
2019 ◽  
Vol 8 (3) ◽  
pp. 256
Author(s):  
Francesco Ponzio ◽  
Gianvito Urgese ◽  
Elisa Ficarra ◽  
Santa Di Cataldo
Keyword(s):  
Neural Networks ◽  
Transfer Learning ◽  
Convolutional Neural Networks ◽  
Digital Pathology ◽  
Training Data ◽  
Automated Classification ◽  
Computationally Efficient ◽  
Deep Convolutional Neural Networks ◽  
Cad Systems ◽  
Tissue Characteristics

Thanks to their capability to learn generalizable descriptors directly from images, deep Convolutional Neural Networks (CNNs) seem the ideal solution to most pattern recognition problems. On the other hand, to learn the image representation, CNNs need huge sets of annotated samples that are unfeasible in many every-day scenarios. This is the case, for example, of Computer-Aided Diagnosis (CAD) systems for digital pathology, where additional challenges are posed by the high variability of the cancerous tissue characteristics. In our experiments, state-of-the-art CNNs trained from scratch on histological images were less accurate and less robust to variability than a traditional machine learning framework, highlighting all the issues of fully training deep networks with limited data from real patients. To solve this problem, we designed and compared three transfer learning frameworks, leveraging CNNs pre-trained on non-medical images. This approach obtained very high accuracy, requiring much less computational resource for the training. Our findings demonstrate that transfer learning is a solution to the automated classification of histological samples and solves the problem of designing accurate and computationally-efficient CAD systems with limited training data.

scholarly journals Thermal-based early breast cancer detection using inception V3, inception V4 and modified inception MV4

2021 ◽  
Author(s):  
Mohammed Abdulla Salim Al Husaini ◽  
Mohamed Hadi Habaebi ◽  
Teddy Surya Gunawan ◽  
Md Rafiqul Islam ◽  
Elfatih A. A. Elsheikh ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Neural Networks ◽  
Deep Learning ◽  
Convolutional Neural Networks ◽  
Early Stage ◽  
Computational Cost ◽  
Optimization Methods ◽  
Optimization Method ◽  
Deep Convolutional Neural Networks ◽  
Breast Thermography

AbstractBreast cancer is one of the most significant causes of death for women around the world. Breast thermography supported by deep convolutional neural networks is expected to contribute significantly to early detection and facilitate treatment at an early stage. The goal of this study is to investigate the behavior of different recent deep learning methods for identifying breast disorders. To evaluate our proposal, we built classifiers based on deep convolutional neural networks modelling inception V3, inception V4, and a modified version of the latter called inception MV4. MV4 was introduced to maintain the computational cost across all layers by making the resultant number of features and the number of pixel positions equal. DMR database was used for these deep learning models in classifying thermal images of healthy and sick patients. A set of epochs 3–30 were used in conjunction with learning rates 1 × 10–3, 1 × 10–4 and 1 × 10–5, Minibatch 10 and different optimization methods. The training results showed that inception V4 and MV4 with color images, a learning rate of 1 × 10–4, and SGDM optimization method, reached very high accuracy, verified through several experimental repetitions. With grayscale images, inception V3 outperforms V4 and MV4 by a considerable accuracy margin, for any optimization methods. In fact, the inception V3 (grayscale) performance is almost comparable to inception V4 and MV4 (color) performance but only after 20–30 epochs. inception MV4 achieved 7% faster classification response time compared to V4. The use of MV4 model is found to contribute to saving energy consumed and fluidity in arithmetic operations for the graphic processor. The results also indicate that increasing the number of layers may not necessarily be useful in improving the performance.

scholarly journals Pashtu Numerals Recognition through Convolutional Neural Networks

2019 ◽  
pp. 91-96
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Convolutional Neural Networks ◽  
Character Recognition ◽  
Optical Character Recognition ◽  
Recognition Accuracy ◽  
Research Use ◽  
Optical Character ◽  
Classification Tasks ◽  
Scanned Images

In the proposed paper we introduce a new Pashtu numerals dataset having handwritten scanned images. We make the dataset publically available for scientific and research use. Pashtu language is used by more than fifty million people both for oral and written communication, but still no efforts are devoted to the Optical Character Recognition (OCR) system for Pashtu language. We introduce a new method for handwritten numerals recognition of Pashtu language through the deep learning based models. We use convolutional neural networks (CNNs) both for features extraction and classification tasks. We assess the performance of the proposed CNNs based model and obtained recognition accuracy of 91.45%.

Sign in / Sign up

Export Citation Format

Share Document