• Nutrition Research and Practice
• /
• v.13 no.6
• /
• pp.521-528
• /
• 2019

BACKGROUND/OBJECTIVES: The aim of this study was to develop Korean food image detection and recognition model for use in mobile devices for accurate estimation of dietary intake. MATERIALS/METHODS: We collected food images by taking pictures or by searching web images and built an image dataset for use in training a complex recognition model for Korean food. Augmentation techniques were performed in order to increase the dataset size. The dataset for training contained more than 92,000 images categorized into 23 groups of Korean food. All images were down-sampled to a fixed resolution of $150{\times}150$ and then randomly divided into training and testing groups at a ratio of 3:1, resulting in 69,000 training images and 23,000 test images. We used a Deep Convolutional Neural Network (DCNN) for the complex recognition model and compared the results with those of other networks: AlexNet, GoogLeNet, Very Deep Convolutional Neural Network, VGG and ResNet, for large-scale image recognition. RESULTS: Our complex food recognition model, K-foodNet, had higher test accuracy (91.3%) and faster recognition time (0.4 ms) than those of the other networks. CONCLUSION: The results showed that K-foodNet achieved better performance in detecting and recognizing Korean food compared to other state-of-the-art models.

• The Journal of the Acoustical Society of Korea
• /
• v.40 no.5
• /
• pp.439-451
• /
• 2021

We propose a deep learning-based beamformer combined with spectral subtraction for continuous speech recognition operating in noisy environments. Conventional beamforming systems were mostly evaluated by using pre-segmented audio signals which were typically generated by mixing speech and noise continuously on a computer. However, since speech utterances are sparsely uttered along the time axis in real environments, conventional beamforming systems degrade in case when noise-only signals without speech are input. To alleviate this drawback, we combine online beamforming algorithm and spectral subtraction. We construct a Continuous Speech Enhancement (CSE) evaluation set to evaluate the online beamforming algorithm in noisy environments. The evaluation set is built by mixing sparsely-occurring speech utterances of the CHiME3 evaluation set and continuously-played CHiME3 background noise and background music of MUSDB. Using a Kaldi-based toolkit and Google web speech recognizer as a speech recognition back-end, we confirm that the proposed online beamforming algorithm with spectral subtraction shows better performance than the baseline online algorithm.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.11
• /
• pp.1403-1410
• /
• 2020

Recently, as interest in non-face-to-face experiences and services increases, the demand for web video contents that can be easily consumed using mobile devices such as smartphones or tablets is rapidly increasing. To cope with these requirements, in this paper we propose a technique to efficiently produce video contents that can provide experience of visiting famous places (i.e., stage tour) in animation or movies. To this end, an image dataset was established by collecting images of stage areas using Google Maps and Google Street View APIs. Afterwards, a deep learning-based style transfer method to apply the unique style of animation videos to the collected street view images and generate the video contents from the style-transferred images was presented. Finally, we showed that the proposed method could produce more interesting stage-tour video contents through various experiments.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.4
• /
• pp.467-473
• /
• 2020

Recently, the number of traffic accidents has also increased with the increase in the penetration rate of cars in Korea. In particular, not only inter-vehicle accidents but also human accidents near crosswalks are increasing, so that more attention to traffic safety around crosswalks are required. In this paper, we propose a system for predicting the safety level around the crosswalk by recognizing an approaching vehicle and estimating the speed of the vehicle using NVIDIA Jetson Nano-class edge devices. To this end, various machine learning models are trained with the information obtained from deep learning-based vehicle detection to predict the degree of risk according to the speed of an approaching vehicle. Finally, based on experiments using actual driving images and web simulation, the performance and the feasibility of the proposed system are validated.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.7
• /
• pp.949-955
• /
• 2022

Deep learning-based text summarization models are not free from datasets. For example, a summarization model trained with a news summarization dataset is not good at summarizing other types of texts such as internet posts and papers. In this study, we define this phenomenon as Data Bias Problem (DBP) and propose two training methods for solving it. The first is the 'proper nouns masking' that masks proper nouns. The second is the 'length variation' that randomly inflates or deflates the length of text. As a result, experiments show that our methods are efficient for solving DBP. In addition, we analyze the results of the experiments and present future development directions. Our contributions are as follows: (1) We discovered DBP and defined it for the first time. (2) We proposed two efficient training methods and conducted actual experiments. (3) Our methods can be applied to all summarization models and are easy to implement, so highly practical.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.23 no.3
• /
• pp.193-198
• /
• 2023

We study a learner-customized lecture recommendation project using deep learning. Recommendation systems can be easily found on the web and apps, and examples using this feature include recommending feature videos by clicking users and advertising items in areas of interest to users on SNS. In this study, the sentence similarity Word2Vec was mainly used to filter twice, and the course was recommended through the Surprise library. With this system, it provides users with the desired classification of course data conveniently and conveniently. Surprise Library is a Python scikit-learn-based library that is conveniently used in recommendation systems. By analyzing the data, the system is implemented at a high speed, and deeper learning is used to implement more precise results through course steps. When a user enters a keyword of interest, similarity between the keyword and the course title is executed, and similarity with the extracted video data and voice text is executed, and the highest ranking video data is recommended through the Surprise Library.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.465-467
• /
• 2022

Mobile phones have become an essential item nowadays since it provides access to online platform and service fast and easy. Coming to these platforms such as Social Network Service (SNS) for shopping have been a go-to option for many people. However, searching for a specific fashion item in the picture is challenging, where users need to try multiple searches by combining appropriate search keywords. To tackle this problem, we propose a system that could provide immediate access to websites related to fashion items. In the framework, we also propose a deep learning model for an automatic analysis of image contexts using instance segmentation. We use transfer learning by utilizing Deep fashion 2 to maximize our model accuracy. After segmenting all the fashion item objects in the image, the related search information is retrieved when the object is clicked. Furthermore, we successfully deploy our system so that it could be assessable using any web browser. We prove that deep learning could be a promising tool not only for scientific purpose but also applicable to commercial shopping.

• Economic and Environmental Geology
• /
• v.46 no.4
• /
• pp.351-358
• /
• 2013

Methane burns more clearly than any other fossil fuels. Coalbed methane(CBM) is natural gas contained in coal beds. This gas is usually producted from coal that is either too deep or too poor-quality to be mined commercially. While global coalbed methane resource estimates are rough, they indicate between 84 and 377tcm, which compares with proven natural gas reserves of 180tcm. Coalbed methane resources are currently only produced on a major scale in the United States, Canada, Australia and China. In this study, we analysed total 109 published papers for the CBM during the 1990~2012 periods by the programs of 'web of science'. The results of analysis, the CBM study led by the United States, the follow India and Australia. In subject area(web of sciences), Energy Fuels is 57, Engineering 58 and Geology 41 papers, respectively.

• Steel and Composite Structures
• /
• v.9 no.5
• /
• pp.419-444
• /
• 2009

Flange and web local buckling in beam plastic hinge regions of steel moment frames can prevent beam-column connections from achieving adequate plastic rotations under earthquake-induced forces. Reducing the flange-web slenderness ratios (FSR/WSR) of beams is the most effective way in mitigating local member buckling as stipulated in the latest seismic design specifications. However, existing steel moment frame buildings with beams that lack the adequate slenderness ratios set forth for new buildings are vulnerable to local member buckling and thereby system-wise instability prior to reaching the required plastic rotation capacities specified for new buildings. This paper presents results from a research study investigating the cyclic behavior of steel I-beams modified by a welded haunch at the bottom flange and reinforced with glass fiber reinforced polymers at the plastic hinge region. Cantilever I-sections with a triangular haunch at the bottom flange and flange slenderness ratios higher then those stipulated in current design specifications were analyzed under reversed cyclic loading. Beam sections with different depth/width and flange/web slenderness ratios (FSR/WSR) were considered. The effect of GFRP thickness, width, and length on stabilizing plastic local buckling was investigated. The FEA results revealed that the contribution of GFRP strips to mitigation of local buckling increases with increasing depth/width ratio and decreasing FSR and WSR. Provided that the interfacial shear strength of the steel/GFRP bond surface is at least 15 MPa, GFRP reinforcement can enable deep beams with FSR of 8-9 and WSR below 55 to maintain plastic rotations in the order of 0.02 radians without experiencing any local buckling.

• Journal of Practical Engineering Education
• /
• v.14 no.2
• /
• pp.351-359
• /
• 2022

In this paper, this app is an online learning tool web application that helps learners learn efficiently. It discusses how learners can improve their learning efficiency in these three aspects: retrieval practice, systematization, metacognition. Through this web service, learners can proceed with learning with a flash card-based retrieval practice. In this case, a method of managing a flash card in a form similar to a directory-file system using a composite pattern is described. Learners can systematically organize their knowledge by converting flash cards into a mind map. The color of the mind map varies according to the learner's learning progress, and learners can easily recognize what they know and what they do not know through color. In this case, it is proposed to build a deep learning model to improve the accuracy of an algorithm for determining and predicting learning progress.