• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.6
• /
• pp.1833-1848
• /
• 2022

Named entity recognition (NER) is an important basic task in the field of Natural Language Processing (NLP). Recently deep learning approaches by extracting word segmentation or character features have been proved to be effective for Chinese Named Entity Recognition (CNER). However, since this method of extracting features only focuses on extracting some of the features, it lacks textual information mining from multiple perspectives and dimensions, resulting in the model not being able to fully capture semantic features. To tackle this problem, we propose a novel Multi-view Semantic Feature Fusion Model (MSFM). The proposed model mainly consists of two core components, that is, Multi-view Semantic Feature Fusion Embedding Module (MFEM) and Multi-head Self-Attention Mechanism Module (MSAM). Specifically, the MFEM extracts character features, word boundary features, radical features, and pinyin features of Chinese characters. The acquired font shape, font sound, and font meaning features are fused to enhance the semantic information of Chinese characters with different granularities. Moreover, the MSAM is used to capture the dependencies between characters in a multi-dimensional subspace to better understand the semantic features of the context. Extensive experimental results on four benchmark datasets show that our method improves the overall performance of the CNER model.

• Communications for Statistical Applications and Methods
• /
• v.26 no.6
• /
• pp.591-610
• /
• 2019

Deep Learning is one of the machine learning methods to find features from a huge data using non-linear transformation. It is now commonly used for supervised learning in many fields. In particular, Convolutional Neural Network (CNN) is the best technique for the image classification since 2012. For users who consider deep learning models for real-world applications, Keras is a popular API for neural networks written in Python and also can be used in R. We try examine the parameter estimation procedures of Deep Neural Network and structures of CNN models from basics to advanced techniques. We also try to figure out some crucial steps in CNN that can improve image classification performance in the CIFAR10 dataset using Keras. We found that several stacks of convolutional layers and batch normalization could improve prediction performance. We also compared image classification performances with other machine learning methods, including K-Nearest Neighbors (K-NN), Random Forest, and XGBoost, in both MNIST and CIFAR10 dataset.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.8
• /
• pp.3790-3803
• /
• 2018

Content-based image retrieval is an approach used to query images based on their semantics. Semantic based retrieval has its application in all fields including medicine, space, computing etc. Semantically generated binary hash codes can improve content-based image retrieval. These semantic labels / binary hash codes can be generated from unlabeled data using convolutional autoencoders. Proposed approach uses semi-supervised deep hashing with semantic learning and binary code generation by minimizing the objective function. Convolutional autoencoders are basis to extract semantic features due to its property of image generation from low level semantic representations. These representations of images are more effective than simple feature extraction and can preserve better semantic information. Proposed activation and loss functions helped to minimize classification error and produce better hash codes. Most widely used datasets have been used for verification of this approach that outperforms the existing methods.

• ETRI Journal
• /
• v.40 no.5
• /
• pp.643-652
• /
• 2018

Dysarthria is a degenerative disorder of the central nervous system that affects the control of articulation and pitch; therefore, it affects the uniqueness of sound produced by the speaker. Hence, dysarthric speaker recognition is a challenging task. In this paper, a feature-extraction method based on deep belief networks is presented for the task of identifying a speaker suffering from dysarthria. The effectiveness of the proposed method is demonstrated and compared with well-known Mel-frequency cepstral coefficient features. For classification purposes, the use of a multi-layer perceptron neural network is proposed with two structures. Our evaluations using the universal access speech database produced promising results and outperformed other baseline methods. In addition, speaker identification under both text-dependent and text-independent conditions are explored. The highest accuracy achieved using the proposed system is 97.3%.

• ETRI Journal
• /
• v.43 no.5
• /
• pp.881-890
• /
• 2021

Photoplethysmography (PPG) is a noninvasive technique that can be used to conveniently measure heart rate (HR) and thus obtain relevant health-related information. However, developing an automated PPG system is difficult, because its waveforms are susceptible to motion artifacts and between-patient variation, making its interpretation difficult. We use deep neural network (DNN) filters to mimic the cognitive ability of a human expert who can distinguish the features of PPG altered by noise from various sources. Systolic (S), onset (O), and first derivative peaks (W) are recognized by three different DNN filters. In addition, the boundaries of uninformative regions caused by artifacts are identified by two different filters. The algorithm reliably derives the HR and presents recognition scores for the S, O, and W peaks and artifacts with only a 0.7-s delay. In the evaluation using data from 11 patients obtained from PhysioNet, the algorithm yields 8643 (86.12%) reliable HR measurements from a total of 10 036 heartbeats, including some with uninformative data resulting from arrhythmias and artifacts.

• International Journal of Advanced Culture Technology
• /
• v.7 no.2
• /
• pp.195-200
• /
• 2019

Media art, which is a combination of media technology and art, is making a lot of progress in combination with AI, IoT and VR. This paper aims to meet people's needs by creating a video that simulates the dance moves of an object that users admire by using media art that features interactive interactions between users and works. The project proposed a universal image synthesis system that minimizes equipment constraints by utilizing a deep running-based Skeleton estimation system and one of the deep-running neural network structures, rather than a Kinect-based Skeleton image. The results of the experiment showed that the images implemented through the deep learning system were successful in generating the same results as the user did when they actually danced through inference and synthesis of motion that they did not actually behave.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2019.05a
• /
• pp.549-550
• /
• 2019

In this paper, we propose a method to detect the scene change using deep learning. To extract feature points, we use a deep neural network and express extracted feature points as 128 dimensional vectors using SIFT descriptor. If it is less than 25%, it is determined that the scene is changed.

• The Journal of Korea Robotics Society
• /
• v.16 no.2
• /
• pp.130-136
• /
• 2021

In recent years, much study has been conducted in robotic grasping. The grasping algorithms based on deep learning have shown better grasping performance than the traditional ones. However, deep learning-based algorithms require a lot of data and time for training. In this study, a grasping algorithm using an artificial neural network-based graspability estimator is proposed. This graspability estimator can be trained with a small number of data by using a neural network based on the residual blocks and point clouds containing the shapes of objects, not RGB images containing various features. The trained graspability estimator can measures graspability of objects and choose the best one to grasp. It was experimentally shown that the proposed algorithm has a success rate of 90% and a cycle time of 12 sec for one grasp, which indicates that it is an efficient grasping algorithm.

• Journal of Korea Multimedia Society
• /
• v.23 no.12
• /
• pp.1447-1453
• /
• 2020

In this paper, we propose a method for no-reference quality assessment of sports videos using 3D shearlet transform and deep residual neural networks. In the proposed method, 3D shearlet transform-based spatiotemporal features are extracted from the overlapped video blocks and applied to logistic regression concatenated with a deep residual neural network based on a conditional video block-wise constraint to learn the spatiotemporal correlation and predict the quality score. Our evaluation reveals that the proposed method predicts the video quality with higher accuracy than the conventional no-reference video quality assessment methods.

• Journal of Sensor Science and Technology
• /
• v.30 no.2
• /
• pp.76-81
• /
• 2021

In this study, we propose a machine vision system with a high object recognition rate. By utilizing a multiple-exposure image sensing technique, the proposed deep learning-based machine vision system can cover a wide light intensity range without further learning processes on the various light intensity range. If the proposed machine vision system fails to recognize object features, the system operates in a multiple-exposure sensing mode and detects the target object that is blocked in the near dark or bright region. Furthermore, short- and long-exposure images from the multiple-exposure sensing mode are synthesized to obtain accurate object feature information. That results in the generation of a wide dynamic range of image information. Even with the object recognition resources for the deep learning process with a light intensity range of only 23 dB, the prototype machine vision system with the multiple-exposure imaging method demonstrated an object recognition performance with a light intensity range of up to 96 dB.