A Novel Memory-Scheduling Strategy for Large Convolutional Neural Network on Memory-Limited Devices

Recently, machine learning, especially deep learning, has been a core algorithm to be widely used in many fields such as natural language processing, speech recognition, object recognition, and so on. At the same time, another trend is that more and more applications are moved to wearable and mobile devices. However, traditional deep learning methods such as convolutional neural network (CNN) and its variants consume a lot of memory resources. In this case, these powerful deep learning methods are difficult to apply on mobile memory-limited platforms. In order to solve this problem, we present a novel memory-management strategy called mmCNN in this paper. With the help of this method, we can easily deploy a trained large-size CNN on any memory size platform such as GPU, FPGA, or memory-limited mobile devices. In our experiments, we run a feed-forward CNN process in some extremely small memory sizes (as low as 5 MB) on a GPU platform. The result shows that our method saves more than 98% memory compared to a traditional CNN algorithm and further saves more than 90% compared to the state-of-the-art related work “vDNNs” (virtualized deep neural networks). Our work in this paper improves the computing scalability of lightweight applications and breaks the memory bottleneck of using deep learning method on memory-limited devices.

Download Full-text

A Review of Plant Phenotypic Image Recognition Technology Based on Deep Learning

Electronics ◽

10.3390/electronics10010081 ◽

2021 ◽

Vol 10 (1) ◽

pp. 81

Author(s):

Jianbin Xiong ◽

Dezheng Yu ◽

Shuangyin Liu ◽

Lei Shu ◽

Xiaochan Wang ◽

...

Keyword(s):

Neural Network ◽

Deep Learning ◽

Convolutional Neural Network ◽

Plant Species ◽

Image Recognition ◽

Recurrent Neural Network ◽

Plant Diseases ◽

Learning Methods ◽

Smart Agriculture ◽

Important Branch

Plant phenotypic image recognition (PPIR) is an important branch of smart agriculture. In recent years, deep learning has achieved significant breakthroughs in image recognition. Consequently, PPIR technology that is based on deep learning is becoming increasingly popular. First, this paper introduces the development and application of PPIR technology, followed by its classification and analysis. Second, it presents the theory of four types of deep learning methods and their applications in PPIR. These methods include the convolutional neural network, deep belief network, recurrent neural network, and stacked autoencoder, and they are applied to identify plant species, diagnose plant diseases, etc. Finally, the difficulties and challenges of deep learning in PPIR are discussed.

Download Full-text

Validating Deep Neural Networks for Online Decoding of Motor Imagery Movements from EEG Signals

Sensors ◽

10.3390/s19010210 ◽

2019 ◽

Vol 19 (1) ◽

pp. 210 ◽

Cited By ~ 32

Author(s):

Zied Tayeb ◽

Juri Fedjaev ◽

Nejla Ghaboosi ◽

Christoph Richter ◽

Lukas Everding ◽

...

Keyword(s):

Neural Network ◽

Machine Learning ◽

Deep Learning ◽

Convolutional Neural Network ◽

Motor Imagery ◽

Classification Performance ◽

Feature Engineering ◽

Learning Models ◽

Eeg Signals ◽

Learning Methods

Non-invasive, electroencephalography (EEG)-based brain-computer interfaces (BCIs) on motor imagery movements translate the subject’s motor intention into control signals through classifying the EEG patterns caused by different imagination tasks, e.g., hand movements. This type of BCI has been widely studied and used as an alternative mode of communication and environmental control for disabled patients, such as those suffering from a brainstem stroke or a spinal cord injury (SCI). Notwithstanding the success of traditional machine learning methods in classifying EEG signals, these methods still rely on hand-crafted features. The extraction of such features is a difficult task due to the high non-stationarity of EEG signals, which is a major cause by the stagnating progress in classification performance. Remarkable advances in deep learning methods allow end-to-end learning without any feature engineering, which could benefit BCI motor imagery applications. We developed three deep learning models: (1) A long short-term memory (LSTM); (2) a spectrogram-based convolutional neural network model (CNN); and (3) a recurrent convolutional neural network (RCNN), for decoding motor imagery movements directly from raw EEG signals without (any manual) feature engineering. Results were evaluated on our own publicly available, EEG data collected from 20 subjects and on an existing dataset known as 2b EEG dataset from “BCI Competition IV”. Overall, better classification performance was achieved with deep learning models compared to state-of-the art machine learning techniques, which could chart a route ahead for developing new robust techniques for EEG signal decoding. We underpin this point by demonstrating the successful real-time control of a robotic arm using our CNN based BCI.

Download Full-text

Dynamic Resource Allocation and Memory Management using Deep Convolutional Neural Network

International Journal of Engineering and Advanced Technology - Regular Issue ◽

10.35940/ijeat.a9961.129219 ◽

2019 ◽

Vol 9 (2) ◽

pp. 608-612

Keyword(s):

Neural Network ◽

Resource Allocation ◽

Deep Learning ◽

Convolutional Neural Network ◽

Data Storage ◽

Memory Management ◽

Research Work ◽

Dynamic Resource Allocation ◽

Mobile Platform ◽

Dynamic Resource

Memory management is very essential task for large-scale storage systems; in mobile platform generate storage errors due to insufficient memory as well as additional task overhead. Many existing systems have illustrated different solution for such issues, like load balancing and load rebalancing. Different unusable applications which are already installed in mobile platform user never access frequently but it allocates some memory space on hard device storage. In the proposed research work we describe dynamic resource allocation for mobile platforms using deep learning approach. In Real world mobile systems users may install different kind of applications which required ad-hoc basis. Such applications may be affect to execution performance of system as well space complexity, sometime they also affect another runnable applications performance. To eliminate of such issues, we carried out an approach to allocate runtime resources for data storage for mobile platform. When system connected with cloud data server it store complete file system on remote Virtual Machine (VM) and whenever a single application required which immediately install beginning as remote server to local device. For developed of proposed system we implemented deep learning base Convolutional Neural Network (CNN), algorithm has used with tensorflow environment which reduces the time complexity for data storage as well as extraction respectively.

Download Full-text

Exploring the Effectiveness of Convolutional Neural Network with Ensemble Technique

10.21203/rs.2.21664/v1 ◽

2020 ◽

Author(s):

Monalisha Ghosh ◽

Goutam Sanyal

Keyword(s):

Neural Network ◽

Machine Learning ◽

Deep Learning ◽

Convolutional Neural Network ◽

Language Processing ◽

Research Work ◽

Research Field ◽

Svm Classifier ◽

Binary Particle Swarm Optimization ◽

Ensemble Technique

Abstract Sentiment Analysis has recently been considered as the most active research field in the natural language processing (NLP) domain. Deep Learning is a subset of the large family of Machine Learning and becoming a growing trend due to its automatic learning capability with impressive consequences across different NLP tasks. Hence, a fusion-based Machine Learning framework has been attempted by merging the Traditional Machine Learning method with Deep Learning techniques to tackle the challenge of sentiment prediction for a massive amount of unstructured review dataset. The proposed architecture aims to utilize the Convolutional Neural Network (CNN) with a backpropagation algorithm to extract embedded feature vectors from the top hidden layer. Thereafter, these vectors augmented to an optimized feature set generated from binary particle swarm optimization (BPSO) method. Finally, a traditional SVM classifier is trained with these extended features set to determine the optimal hyper-plane for separating two classes of review datasets. The evaluation of this research work has been carried out on two benchmark movie review datasets IMDB, SST2. Experimental results with comparative studies based on performance accuracy and F-score value are reported to highlight the benefits of the developed frameworks.

Download Full-text

Application of deep learning models for aspect based sentiment analysis.

Bulletin of Bryansk state technical university ◽

10.12737/22917 ◽

2015 ◽

Vol 2015 (3) ◽

pp. 117-126

Author(s):

Дмитрий Будыльский ◽

Dmitriy Budylskiy ◽

Александр Подвесовский ◽

Aleksandr Podvesovskiy

Keyword(s):

Neural Network ◽

Deep Learning ◽

Natural Language Processing ◽

Convolutional Neural Network ◽

Language Processing ◽

Recurrent Neural Network ◽

Russian Text ◽

Learning Models ◽

Actual Problem ◽

Good Efficiency

This paper describes actual problem of sentiment based aspect analysis and four deep learning models: convolutional neural network, recurrent neural network, GRU and LSTM networks. We evaluated these models on Russian text dataset from SentiRuEval-2015. Results show good efficiency and high potential for further natural language processing applications.

Download Full-text

A Bi-Directional LSTM-CNN Model with Attention for Aspect-Level Text Classification

Future Internet ◽

10.3390/fi10120116 ◽

2018 ◽

Vol 10 (12) ◽

pp. 116 ◽

Cited By ~ 9

Author(s):

Yonghua Zhu ◽

Xun Gao ◽

Weilin Zhang ◽

Shenkai Liu ◽

Yuanyuan Zhang

Keyword(s):

Neural Network ◽

Deep Learning ◽

Convolutional Neural Network ◽

Text Classification ◽

Business Value ◽

The Internet ◽

Information Analysis ◽

Target Information ◽

Learning Methods ◽

Daily Lives

The prevalence that people share their opinions on the products and services in their daily lives on the Internet has generated a large quantity of comment data, which contain great business value. As for comment sentences, they often contain several comment aspects and the sentiment on these aspects are different, which makes it meaningless to give an overall sentiment polarity of the sentence. In this paper, we introduce Attention-based Aspect-level Recurrent Convolutional Neural Network (AARCNN) to analyze the remarks at aspect-level. The model integrates attention mechanism and target information analysis, which enables the model to concentrate on the important parts of the sentence and to make full use of the target information. The model uses bidirectional LSTM (Bi-LSTM) to build the memory of the sentence, and then CNN is applied to extracting attention from memory to get the attentive sentence representation. The model uses aspect embedding to analyze the target information of the representation and finally the model outputs the sentiment polarity through a softmax layer. The model was tested on multi-language datasets, and demonstrated that it has better performance than conventional deep learning methods.

Download Full-text

A review on convolutional neural network based deep learning methods in gene expression data for disease diagnosis

Materials Today Proceedings ◽

10.1016/j.matpr.2020.10.263 ◽

2020 ◽

Author(s):

C. Gunavathi ◽

K. Sivasubramanian ◽

P. Keerthika ◽

C. Paramasivam

Keyword(s):

Neural Network ◽

Gene Expression ◽

Deep Learning ◽

Convolutional Neural Network ◽

Gene Expression Data ◽

Disease Diagnosis ◽

Expression Data ◽

Learning Methods

Download Full-text

Identifying Emotion on Indonesian Tweets using Convolutional Neural Networks

Jurnal RESTI (Rekayasa Sistem dan Teknologi Informasi) ◽

10.29207/resti.v5i3.3137 ◽

2021 ◽

Vol 5 (3) ◽

pp. 584-593

Author(s):

Naufal Hilmiaji ◽

Kemas Muslim Lhaksmana ◽

Mahendra Dwifebri Purbolaksono

Keyword(s):

Neural Network ◽

Neural Networks ◽

Performance Evaluation ◽

Deep Learning ◽

Convolutional Neural Network ◽

Convolutional Neural Networks ◽

Text Classification ◽

Classification Model ◽

Learning Methods ◽

Expected Performance

especially with the advancement of deep learning methods for text classification. Despite some effort to identify emotion on Indonesian tweets, its performance evaluation results have not achieved acceptable numbers. To solve this problem, this paper implements a classification model using a convolutional neural network (CNN), which has demonstrated expected performance in text classification. To easily compare with the previous research, this classification is performed on the same dataset, which consists of 4,403 tweets in Indonesian that were labeled using five different emotion classes: anger, fear, joy, love, and sadness. The performance evaluation results achieve the precision, recall, and F1-score at respectively 90.1%, 90.3%, and 90.2%, while the highest accuracy achieves 89.8%. These results outperform previous research that classifies the same classification on the same dataset.

Download Full-text

A Spectral-Spatial Cascaded 3D Convolutional Neural Network with a Convolutional Long Short-Term Memory Network for Hyperspectral Image Classification

Remote Sensing ◽

10.3390/rs11202363 ◽

2019 ◽

Vol 11 (20) ◽

pp. 2363 ◽

Cited By ~ 2

Author(s):

Wenchao Qi ◽

Xia Zhang ◽

Nan Wang ◽

Mao Zhang ◽

Yi Cen

Keyword(s):

Neural Network ◽

Deep Learning ◽

Convolutional Neural Network ◽

Short Term Memory ◽

Hyperspectral Image ◽

Short Term ◽

Dynamic Learning ◽

Learning Methods ◽

Term Memory ◽

Long Short Term Memory

Deep learning methods used for hyperspectral image (HSI) classification often achieve greater accuracy than traditional algorithms but require large numbers of training epochs. To simplify model structures and reduce their training epochs, an end-to-end deep learning framework incorporating a spectral-spatial cascaded 3D convolutional neural network (CNN) with a convolutional long short-term memory (CLSTM) network, called SSCC, is proposed herein for HSI classification. The SSCC framework employs cascaded 3D CNN to learn the spectral-spatial features of HSIs and uses the CLSTM network to extract sequence features. Residual connections are used in SSCC to accelerate model convergence, with the outputs of previous convolutional layers concatenated as inputs for subsequent layers. Moreover, the data augmentation, parametric rectified linear unit, dynamic learning rate, batch normalization, and regularization (including dropout and L2) methods are used to increase classification accuracy and prevent overfitting. These attributes allow the SSCC framework to achieve good performance for HSI classification within 20 epochs. Three well-known datasets including Indiana Pines, University of Pavia, and Pavia Center were employed to evaluate the classification performance of the proposed algorithm. The GF-5 dataset of Anxin County, obtained from China’s recently launched spaceborne Advanced Hyperspectral Imager, was also used for classification experiments. The experimental results demonstrate that the proposed SSCC framework achieves state-of-the-art performance with better training efficiency than other deep learning methods.

Download Full-text

Identification Method of Wheat Cultivars by Using a Convolutional Neural Network Combined with Images of Multiple Growth Periods of Wheat

Symmetry ◽

10.3390/sym13112012 ◽

2021 ◽

Vol 13 (11) ◽

pp. 2012

Author(s):

Jiameng Gao ◽

Chengzhong Liu ◽

Junying Han ◽

Qinglin Lu ◽

Hengxing Wang ◽

...

Keyword(s):

Neural Network ◽

Machine Learning ◽

Deep Learning ◽

Convolutional Neural Network ◽

Wheat Variety ◽

Learning Model ◽

Wheat Cultivars ◽

Learning Methods ◽

Wheat Varieties ◽

Deep Learning Model

Wheat is a very important food crop for mankind. Many new varieties are bred every year. The accurate judgment of wheat varieties can promote the development of the wheat industry and the protection of breeding property rights. Although gene analysis technology can be used to accurately determine wheat varieties, it is costly, time-consuming, and inconvenient. Traditional machine learning methods can significantly reduce the cost and time of wheat cultivars identification, but the accuracy is not high. In recent years, the relatively popular deep learning methods have further improved the accuracy on the basis of traditional machine learning, whereas it is quite difficult to continue to improve the identification accuracy after the convergence of the deep learning model. Based on the ResNet and SENet models, this paper draws on the idea of the bagging-based ensemble estimator algorithm, and proposes a deep learning model for wheat classification, CMPNet, which is coupled with the tillering period, flowering period, and seed image. This convolutional neural network (CNN) model has a symmetrical structure along the direction of the tensor flow. The model uses collected images of different types of wheat in multiple growth periods. First, it uses the transfer learning method of the ResNet-50, SE-ResNet, and SE-ResNeXt models, and then trains the collected images of 30 kinds of wheat in different growth periods. It then uses the concat layer to connect the output layers of the three models, and finally obtains the wheat classification results through the softmax function. The accuracy of wheat variety identification increased from 92.07% at the seed stage, 95.16% at the tillering stage, and 97.38% at the flowering stage to 99.51%. The model’s single inference time was only 0.0212 s. The model not only significantly improves the classification accuracy of wheat varieties, but also achieves low cost and high efficiency, which makes it a novel and important technology reference for wheat producers, managers, and law enforcement supervisors in the practice of wheat production.

Download Full-text