• KIPS Transactions on Computer and Communication Systems
• /
• 제10권10호
• /
• pp.261-268
• /
• 2021

Cloud computing is a computing paradigm in which users can utilize computing resources in a pay-as-you-go manner. In a cloud system, resources can be dynamically scaled up and down to the user's on-demand so that the total cost of ownership can be reduced. The Modeling and Simulation (M&S) technology is a renowned simulation-based method to obtain engineering analysis and results through CAE software without actual experimental action. In general, M&S technology is utilized in Finite Element Analysis (FEA), Computational Fluid Dynamics (CFD), Multibody dynamics (MBD), and optimization fields. The work procedure through M&S is divided into pre-processing, analysis, and post-processing steps. The pre/post-processing are GPU-intensive job that consists of 3D modeling jobs via CAE software, whereas analysis is CPU or GPU intensive. Because a general-purpose desktop needs plenty of time to analyze complicated 3D models, CAE software requires a high-end CPU and GPU-based workstation that can work fluently. In other words, for executing M&S, it is absolutely required to utilize high-performance computing resources. To mitigate the cost issue from equipping such tremendous computing resources, we propose HEMOS-Cloud service, an integrated cloud and cluster computing environment. The HEMOS-Cloud service provides CAE software and computing resources to users who want to experience M&S in business sectors or academics. In this paper, the economic ripple effect of HEMOS-Cloud service was analyzed by using industry-related analysis. The estimated results of using the experts-guided coefficients are the production inducement effect of KRW 7.4 billion, the value-added effect of KRW 4.1 billion, and the employment-inducing effect of 50 persons per KRW 1 billion.

• Journal of Digital Contents Society
• /
• 제16권5호
• /
• pp.805-813
• /
• 2015

Many researchers expect the compressive sensing and sparse recovery problem can overcome the limitation of conventional digital techniques. However, these new approaches require to solve the l1 norm optimization problems when it comes to signal reconstruction. In the signal reconstruction process, the transform computation by multiplication of a random matrix and a vector consumes considerable computing power. To address this issue, parallel processing is applied to the optimization problems. In particular, due to huge size of original signal, it is hard to store the random matrix directly in memory, which makes one need to design a procedural approach in handling the random matrix. This paper presents a new parallel algorithm to calculate random partial Haar wavelet transform based on Single Instruction Multiple Threads (SIMT) platform.

• Journal of Intelligence and Information Systems
• /
• 제26권4호
• /
• pp.27-65
• /
• 2020

Many companies on information and communication technology make public their own developed AI technology, for example, Google's TensorFlow, Facebook's PyTorch, Microsoft's CNTK. By releasing deep learning open source software to the public, the relationship with the developer community and the artificial intelligence (AI) ecosystem can be strengthened, and users can perform experiment, implementation and improvement of it. Accordingly, the field of machine learning is growing rapidly, and developers are using and reproducing various learning algorithms in each field. Although various analysis of open source software has been made, there is a lack of studies to help develop or use deep learning open source software in the industry. This study thus attempts to derive a strategy for adopting the framework through case studies of a deep learning open source framework. Based on the technology-organization-environment (TOE) framework and literature review related to the adoption of open source software, we employed the case study framework that includes technological factors as perceived relative advantage, perceived compatibility, perceived complexity, and perceived trialability, organizational factors as management support and knowledge & expertise, and environmental factors as availability of technology skills and services, and platform long term viability. We conducted a case study analysis of three companies' adoption cases (two cases of success and one case of failure) and revealed that seven out of eight TOE factors and several factors regarding company, team and resource are significant for the adoption of deep learning open source framework. By organizing the case study analysis results, we provided five important success factors for adopting deep learning framework: the knowledge and expertise of developers in the team, hardware (GPU) environment, data enterprise cooperation system, deep learning framework platform, deep learning framework work tool service. In order for an organization to successfully adopt a deep learning open source framework, at the stage of using the framework, first, the hardware (GPU) environment for AI R&D group must support the knowledge and expertise of the developers in the team. Second, it is necessary to support the use of deep learning frameworks by research developers through collecting and managing data inside and outside the company with a data enterprise cooperation system. Third, deep learning research expertise must be supplemented through cooperation with researchers from academic institutions such as universities and research institutes. Satisfying three procedures in the stage of using the deep learning framework, companies will increase the number of deep learning research developers, the ability to use the deep learning framework, and the support of GPU resource. In the proliferation stage of the deep learning framework, fourth, a company makes the deep learning framework platform that improves the research efficiency and effectiveness of the developers, for example, the optimization of the hardware (GPU) environment automatically. Fifth, the deep learning framework tool service team complements the developers' expertise through sharing the information of the external deep learning open source framework community to the in-house community and activating developer retraining and seminars. To implement the identified five success factors, a step-by-step enterprise procedure for adoption of the deep learning framework was proposed: defining the project problem, confirming whether the deep learning methodology is the right method, confirming whether the deep learning framework is the right tool, using the deep learning framework by the enterprise, spreading the framework of the enterprise. The first three steps (i.e. defining the project problem, confirming whether the deep learning methodology is the right method, and confirming whether the deep learning framework is the right tool) are pre-considerations to adopt a deep learning open source framework. After the three pre-considerations steps are clear, next two steps (i.e. using the deep learning framework by the enterprise and spreading the framework of the enterprise) can be processed. In the fourth step, the knowledge and expertise of developers in the team are important in addition to hardware (GPU) environment and data enterprise cooperation system. In final step, five important factors are realized for a successful adoption of the deep learning open source framework. This study provides strategic implications for companies adopting or using deep learning framework according to the needs of each industry and business.

• Journal of Intelligence and Information Systems
• /
• 제28권1호
• /
• pp.89-106
• /
• 2022

Recently, the amount of video data collected from smartphones, CCTVs, black boxes, and high-definition cameras has increased rapidly. According to the increasing video data, the requirements for analysis and utilization are increasing. Due to the lack of skilled manpower to analyze videos in many industries, machine learning and artificial intelligence are actively used to assist manpower. In this situation, the demand for various computer vision technologies such as object detection and tracking, action detection, emotion detection, and Re-ID also increased rapidly. However, the object detection and tracking technology has many difficulties that degrade performance, such as re-appearance after the object's departure from the video recording location, and occlusion. Accordingly, action and emotion detection models based on object detection and tracking models also have difficulties in extracting data for each object. In addition, deep learning architectures consist of various models suffer from performance degradation due to bottlenects and lack of optimization. In this study, we propose an video analysis system consists of YOLOv5 based DeepSORT object tracking model, SlowFast based action recognition model, Torchreid based Re-ID model, and AWS Rekognition which is emotion recognition service. Proposed model uses single-linkage hierarchical clustering based Re-ID and some processing method which maximize hardware throughput. It has higher accuracy than the performance of the re-identification model using simple metrics, near real-time processing performance, and prevents tracking failure due to object departure and re-emergence, occlusion, etc. By continuously linking the action and facial emotion detection results of each object to the same object, it is possible to efficiently analyze videos. The re-identification model extracts a feature vector from the bounding box of object image detected by the object tracking model for each frame, and applies the single-linkage hierarchical clustering from the past frame using the extracted feature vectors to identify the same object that failed to track. Through the above process, it is possible to re-track the same object that has failed to tracking in the case of re-appearance or occlusion after leaving the video location. As a result, action and facial emotion detection results of the newly recognized object due to the tracking fails can be linked to those of the object that appeared in the past. On the other hand, as a way to improve processing performance, we introduce Bounding Box Queue by Object and Feature Queue method that can reduce RAM memory requirements while maximizing GPU memory throughput. Also we introduce the IoF(Intersection over Face) algorithm that allows facial emotion recognized through AWS Rekognition to be linked with object tracking information. The academic significance of this study is that the two-stage re-identification model can have real-time performance even in a high-cost environment that performs action and facial emotion detection according to processing techniques without reducing the accuracy by using simple metrics to achieve real-time performance. The practical implication of this study is that in various industrial fields that require action and facial emotion detection but have many difficulties due to the fails in object tracking can analyze videos effectively through proposed model. Proposed model which has high accuracy of retrace and processing performance can be used in various fields such as intelligent monitoring, observation services and behavioral or psychological analysis services where the integration of tracking information and extracted metadata creates greate industrial and business value. In the future, in order to measure the object tracking performance more precisely, there is a need to conduct an experiment using the MOT Challenge dataset, which is data used by many international conferences. We will investigate the problem that the IoF algorithm cannot solve to develop an additional complementary algorithm. In addition, we plan to conduct additional research to apply this model to various fields' dataset related to intelligent video analysis.

• Journal of Digital Convergence
• /
• 제18권11호
• /
• pp.259-266
• /
• 2020

Many developments have been steadily carried out by researchers with applying knowledge-based expert system or machine learning algorithms to the financial field. In particular, it is now common to perform knowledge based system trading in using stock prices. Recently, deep learning technologies have been applied to real fields of stock trading marketplace as GPU performance and large scaled data have been supported enough. Especially, LSTM has been tried to apply to stock price prediction because of its compatibility for time series data. In this paper, we implement stock price prediction using LSTM. In modeling of LSTM, we propose a fitness combination of model parameters and activation functions for best performance. Specifically, we propose suitable selection methods of initializers of weights and bias, regularizers to avoid over-fitting, activation functions and optimization methods. We also compare model performances according to the different selections of the above important modeling considering factors on the real-world stock price data of global major companies. Finally, our experimental work brings a fitness method of applying LSTM model to stock price prediction.

• Journal of the Korean Society of Radiology
• /
• 제17권5호
• /
• pp.633-640
• /
• 2023

With the recent development of static and dynamic modulated brachytherapy methods in brachytherapy, which use radiation shielding to modulate the dose distribution to deliver the dose, the amount of parameters and data required for dose calculation in inverse treatment planning and treatment plan optimization algorithms suitable for new directional beam intensity modulated brachytherapy is increasing. Although intensity-modulated brachytherapy enables accurate dose delivery of radiation, the increased amount of parameters and data increases the elapsed time required for dose calculation. In this study, a GPU-based CUDA-accelerated dose calculation algorithm was constructed to reduce the increase in dose calculation elapsed time. The acceleration of the calculation process was achieved by parallelizing the calculation of the system matrix of the volume of interest and the dose calculation. The developed algorithms were all performed in the same computing environment with an Intel (3.7 GHz, 6-core) CPU and a single NVIDIA GTX 1080ti graphics card, and the dose calculation time was evaluated by measuring only the dose calculation time, excluding the additional time required for loading data from disk and preprocessing operations. The results showed that the accelerated algorithm reduced the dose calculation time by about 30 times compared to the CPU-only calculation. The accelerated dose calculation algorithm can be expected to speed up treatment planning when new treatment plans need to be created to account for daily variations in applicator movement, such as in adaptive radiotherapy, or when dose calculation needs to account for changing parameters, such as in dynamically modulated brachytherapy.

• Journal of Intelligence and Information Systems
• /
• 제22권2호
• /
• pp.127-142
• /
• 2016

Deep learning model is a kind of neural networks that allows multiple hidden layers. There are various deep learning architectures such as convolutional neural networks, deep belief networks and recurrent neural networks. Those have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition and bioinformatics where they have been shown to produce state-of-the-art results on various tasks. Among those architectures, convolutional neural networks and recurrent neural networks are classified as the supervised learning model. And in recent years, those supervised learning models have gained more popularity than unsupervised learning models such as deep belief networks, because supervised learning models have shown fashionable applications in such fields mentioned above. Deep learning models can be trained with backpropagation algorithm. Backpropagation is an abbreviation for "backward propagation of errors" and a common method of training artificial neural networks used in conjunction with an optimization method such as gradient descent. The method calculates the gradient of an error function with respect to all the weights in the network. The gradient is fed to the optimization method which in turn uses it to update the weights, in an attempt to minimize the error function. Convolutional neural networks use a special architecture which is particularly well-adapted to classify images. Using this architecture makes convolutional networks fast to train. This, in turn, helps us train deep, muti-layer networks, which are very good at classifying images. These days, deep convolutional networks are used in most neural networks for image recognition. Convolutional neural networks use three basic ideas: local receptive fields, shared weights, and pooling. By local receptive fields, we mean that each neuron in the first(or any) hidden layer will be connected to a small region of the input(or previous layer's) neurons. Shared weights mean that we're going to use the same weights and bias for each of the local receptive field. This means that all the neurons in the hidden layer detect exactly the same feature, just at different locations in the input image. In addition to the convolutional layers just described, convolutional neural networks also contain pooling layers. Pooling layers are usually used immediately after convolutional layers. What the pooling layers do is to simplify the information in the output from the convolutional layer. Recent convolutional network architectures have 10 to 20 hidden layers and billions of connections between units. Training deep learning networks has taken weeks several years ago, but thanks to progress in GPU and algorithm enhancement, training time has reduced to several hours. Neural networks with time-varying behavior are known as recurrent neural networks or RNNs. A recurrent neural network is a class of artificial neural network where connections between units form a directed cycle. This creates an internal state of the network which allows it to exhibit dynamic temporal behavior. Unlike feedforward neural networks, RNNs can use their internal memory to process arbitrary sequences of inputs. Early RNN models turned out to be very difficult to train, harder even than deep feedforward networks. The reason is the unstable gradient problem such as vanishing gradient and exploding gradient. The gradient can get smaller and smaller as it is propagated back through layers. This makes learning in early layers extremely slow. The problem actually gets worse in RNNs, since gradients aren't just propagated backward through layers, they're propagated backward through time. If the network runs for a long time, that can make the gradient extremely unstable and hard to learn from. It has been possible to incorporate an idea known as long short-term memory units (LSTMs) into RNNs. LSTMs make it much easier to get good results when training RNNs, and many recent papers make use of LSTMs or related ideas.

• Journal of Intelligence and Information Systems
• /
• 제23권2호
• /
• pp.71-88
• /
• 2017

Language models were originally developed for speech recognition and language processing. Using a set of example sentences, a language model predicts the next word or character based on sequential input data. N-gram models have been widely used but this model cannot model the correlation between the input units efficiently since it is a probabilistic model which are based on the frequency of each unit in the training set. Recently, as the deep learning algorithm has been developed, a recurrent neural network (RNN) model and a long short-term memory (LSTM) model have been widely used for the neural language model (Ahn, 2016; Kim et al., 2016; Lee et al., 2016). These models can reflect dependency between the objects that are entered sequentially into the model (Gers and Schmidhuber, 2001; Mikolov et al., 2010; Sundermeyer et al., 2012). In order to learning the neural language model, texts need to be decomposed into words or morphemes. Since, however, a training set of sentences includes a huge number of words or morphemes in general, the size of dictionary is very large and so it increases model complexity. In addition, word-level or morpheme-level models are able to generate vocabularies only which are contained in the training set. Furthermore, with highly morphological languages such as Turkish, Hungarian, Russian, Finnish or Korean, morpheme analyzers have more chance to cause errors in decomposition process (Lankinen et al., 2016). Therefore, this paper proposes a phoneme-level language model for Korean language based on LSTM models. A phoneme such as a vowel or a consonant is the smallest unit that comprises Korean texts. We construct the language model using three or four LSTM layers. Each model was trained using Stochastic Gradient Algorithm and more advanced optimization algorithms such as Adagrad, RMSprop, Adadelta, Adam, Adamax, and Nadam. Simulation study was done with Old Testament texts using a deep learning package Keras based the Theano. After pre-processing the texts, the dataset included 74 of unique characters including vowels, consonants, and punctuation marks. Then we constructed an input vector with 20 consecutive characters and an output with a following 21st character. Finally, total 1,023,411 sets of input-output vectors were included in the dataset and we divided them into training, validation, testsets with proportion 70:15:15. All the simulation were conducted on a system equipped with an Intel Xeon CPU (16 cores) and a NVIDIA GeForce GTX 1080 GPU. We compared the loss function evaluated for the validation set, the perplexity evaluated for the test set, and the time to be taken for training each model. As a result, all the optimization algorithms but the stochastic gradient algorithm showed similar validation loss and perplexity, which are clearly superior to those of the stochastic gradient algorithm. The stochastic gradient algorithm took the longest time to be trained for both 3- and 4-LSTM models. On average, the 4-LSTM layer model took 69% longer training time than the 3-LSTM layer model. However, the validation loss and perplexity were not improved significantly or became even worse for specific conditions. On the other hand, when comparing the automatically generated sentences, the 4-LSTM layer model tended to generate the sentences which are closer to the natural language than the 3-LSTM model. Although there were slight differences in the completeness of the generated sentences between the models, the sentence generation performance was quite satisfactory in any simulation conditions: they generated only legitimate Korean letters and the use of postposition and the conjugation of verbs were almost perfect in the sense of grammar. The results of this study are expected to be widely used for the processing of Korean language in the field of language processing and speech recognition, which are the basis of artificial intelligence systems.