• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.10a
• /
• pp.958-961
• /
• 2019

Convolutional neural network(CNN)의 등장으로 얼굴 표현 인식 연구는 많은 발전을 이루었다. 그러나, 기존의 CNN 접근법은 미리 학습된 훈련모델에서 Multiple-level 의 의미적 맥락을 포함하지 않는 Attention-embedded 문제가 발생한다. 사람의 얼굴 감정은 다양한 근육의 움직임과 결합에 기초하여 관찰되며, CNN 에서 딥 레이어의 산출물로 나온 특징들의 결합은 많은 서브샘플링 단계를 통해서 class 구별와 같은 의미 정보의 손실이 일어나기 때문에 전이 학습을 통한 올바른 훈련 모델 생성이 어렵다는 단점이 있다. 따라서, 본 논문은 Backbone 네트워크의 Multi-level 특성에서 Channel-wise Attention 통합 및 의미 정보를 포함하여 높은 인식 성능을 달성하는 Densely Backwarnd Attention(DBA) CNN 방법을 제안한다. 제안하는 기법은 High-level 기능에서 채널 간 시멘틱 정보를 활용하여 세분화된 시멘틱 정보를 Low-level 버전에서 다시 재조정한다. 그런 다음, 중요한 얼굴 표정의 묘사를 분명하게 포함시키기 위해서 multi-level 데이터를 통합하는 단계를 추가로 실행한다. 실험을 통해, 제안된 접근방법이 정확도 79.37%를 달성 하여 제안 기술이 효율성이 있음을 증명하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.10
• /
• pp.4080-4097
• /
• 2020

In recent years, convolutional neural network (CNN) has become the primary method for face detection. But its shortcomings are obvious, such as expensive calculation, heavy model, etc. This makes CNN difficult to use on the mobile devices which have limited computing and storage capabilities. Therefore, the design of lightweight CNN for face detection is becoming more and more important with the popularity of smartphones and mobile Internet. Based on the CPU real-time face detector FaceBoxes, we propose a multi-task lightweight face detector, which has low computing cost and higher detection precision. First, to improve the detection capability, the squeeze and excitation modules are used to extract attention between channels. Then, the textual and semantic information are extracted by shallow networks and deep networks respectively to get rich features. Finally, the landmark detection module is used to improve the detection performance for small faces and provide landmark data for face alignment. Experiments on AFW, FDDB, PASCAL, and WIDER FACE datasets show that our algorithm has achieved significant improvement in the mean average precision. Especially, on the WIDER FACE hard validation set, our algorithm outperforms the mean average precision of FaceBoxes by 7.2%. For VGA-resolution images, the running speed of our algorithm can reach 23FPS on a CPU device.

• Journal of IKEEE
• /
• v.22 no.4
• /
• pp.1175-1179
• /
• 2018

In this paper, we propose recurrent CNN(Convolutional Neural Networks) for detecting seizures among patients using EEG signals. In the proposed method, data were mapped by image to preserve the spectral characteristics of the EEG signal and the position of the electrode. After the spectral preprocessing, we input it into CNN and extracted the spatial and temporal features without wavelet transform. Results from the Children's Hospital of Boston Massachusetts Institute of Technology (CHB-MIT) dataset showed a sensitivity of 90% and a false positive rate (FPR) of 0.85 per hour.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.12
• /
• pp.1506-1513
• /
• 2019

Recently, as the industrial scope of multi-rotor unmanned aerial vehicles(UAV) is greatly expanded, the demands for data collection, processing, and analysis using UAV are also increasing. However, the acoustic data collected by using the UAV is greatly corrupted by the UAV's motor noise and wind noise, which makes it difficult to process and analyze the acoustic data. Therefore, we have studied a method to enhance the target sound from the acoustic signal received through microphones connected to UAV. In this paper, we have extended the densely connected dilated convolutional network, one of the existing single channel acoustic enhancement technique, to consider the inter-channel characteristics of the acoustic signal. As a result, the extended model performed better than the existed model in all evaluation measures such as SDR, PESQ, and STOI.

• Journal of the Institute of Convergence Signal Processing
• /
• v.21 no.2
• /
• pp.73-78
• /
• 2020

Musculoskeletal disease is often caused by sitting down for long period's time or by bad posture habits. In order to prevent musculoskeletal disease in daily life, it is the most important to correct the bad sitting posture to the right one through real-time monitoring. In this study, to detect the sitting information of user's without any constraints, we propose posture measurement system based on multi-channel pressure sensor and CNN model for classifying sitting posture types. The proposed CNN model can analyze 5 types of sitting postures based on sitting posture information. For the performance assessment of posture classification CNN model through field test, the accuracy, recall, precision, and F1 of the classification results were checked with 10 subjects. As the experiment results, 99.84% of accuracy, 99.6% of recall, 99.6% of precision, and 99.6% of F1 were verified.

• Journal of Biomedical Engineering Research
• /
• v.43 no.5
• /
• pp.361-368
• /
• 2022

The purpose of this study was to develop a hand rehabilitation training system for hemiplegic patients. We also tried to find out five hand postures (WF: Wrist Flexion, WE: Wrist Extension, BG: Ball Grip, HG: Hook Grip, RE: Rest) in real-time using multi-channel EMG-based deep learning. We performed a pre-processing method that converts to Spider Chart image data for the classification of hand movement from five test subjects (total 1,500 data sets) using Convolution Neural Networks (CNN) deep learning with an 8-channel armband. As a result of this study, the recognition accuracy was 92% for WF, 94% for WE, 76% for BG, 82% for HG, and 88% for RE. Also, ten physical therapists participated for the usability evaluation. The questionnaire consisted of 7 items of acceptance, interest, and satisfaction, and the mean and standard deviation were calculated by dividing each into a 5-point scale. As a result, high scores were obtained in immersion and interest in game (4.6±0.43), convenience of the device (4.9±0.30), and satisfaction after treatment (4.1±0.48). On the other hand, Conformity of intention for treatment (3.90±0.49) was relatively low. This is thought to be because the game play may be difficult depending on the degree of spasticity of the hemiplegic patient, and compensation may occur in patient with weakened target muscles. Therefore, it is necessary to develop a rehabilitation program suitable for the degree of disability of the patient.

• Journal of Positioning, Navigation, and Timing
• /
• v.12 no.1
• /
• pp.1-9
• /
• 2023

The problem of classifying a non-line-of-sight (NLOS) signal in a multipath channel is important to improve global navigation satellite system (GNSS) positioning accuracy in urban areas. Conventional deep learning-based NLOS signal classifiers use GNSS satellite measurements such as the carrier-to-noise-density ratio (CN_0), pseudorange, and elevation angle as inputs. However, there is a computational inefficiency with use of these measurements and the NLOS signal features expressed by the measurements are limited. In this paper, we propose a Convolutional Neural Network (CNN)-based NLOS signal classifier that receives successive Auto-correlation function (ACF) outputs according to a time-series, which is the most primitive output of GNSS signal processing. We compared the proposed classifier to other DL-based NLOS signal classifiers such as a multi-layer perceptron (MLP) and Gated Recurrent Unit (GRU) to show the superiority of the proposed classifier. The results show the proposed classifier does not require the navigation data extraction stage to classify the NLOS signals, and it has been verified that it has the best detection performance among all compared classifiers, with an accuracy of up to 97%.

• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.53-62
• /
• 2022

The deployment of advanced structural health monitoring (SHM) systems in large-scale civil structures collects large amounts of data. Note that these data may contain multiple types of anomalies (e.g., missing, minor, outlier, etc.) caused by harsh environment, sensor faults, transfer omission and other factors. These anomalies seriously affect the evaluation of structural performance. Therefore, the effective analysis and mining of SHM data is an extremely important task. Inspired by the deep learning paradigm, this study develops a novel generative adversarial network (GAN) and convolutional neural network (CNN)-based data anomaly detection approach for SHM. The framework of the proposed approach includes three modules : (a) A three-channel input is established based on fast Fourier transform (FFT) and Gramian angular field (GAF) method; (b) A GANomaly is introduced and trained to extract features from normal samples alone for class-imbalanced problems; (c) Based on the output of GANomaly, a CNN is employed to distinguish the types of anomalies. In addition, a dataset-oriented method (i.e., multistage sampling) is adopted to obtain the optimal sampling ratios between all different samples. The proposed approach is tested with acceleration data from an SHM system of a long-span bridge. The results show that the proposed approach has a higher accuracy in detecting the multi-pattern anomalies of SHM data.

• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.181-193
• /
• 2022

Data-driven structural health monitoring (SHM) of civil infrastructure can be used to continuously assess the state of a structure, allowing preemptive safety measures to be carried out. Long-term monitoring of large-scale civil infrastructure often involves data-collection using a network of numerous sensors of various types. Malfunctioning sensors in the network are common, which can disrupt the condition assessment and even lead to false-negative indications of damage. The overwhelming size of the data collected renders manual approaches to ensure data quality intractable. The task of detecting and classifying an anomaly in the raw data is non-trivial. We propose an approach to automate this task, improving upon the previously developed technique of image-based pre-processing on one-dimensional (1D) data by enriching the features of the neural network input data with multiple channels. In particular, feature engineering is employed to convert the measured time histories into a 3-channel image comprised of (i) the time history, (ii) the spectrogram, and (iii) the probability density function representation of the signal. To demonstrate this approach, a CNN model is designed and trained on a dataset consisting of acceleration records of sensors installed on a long-span bridge, with the goal of fault detection and classification. The effect of imbalance in anomaly patterns observed is studied to better account for unseen test cases. The proposed framework achieves high overall accuracy and recall even when tested on an unseen dataset that is much larger than the samples used for training, offering a viable solution for implementation on full-scale structures where limited labeled-training data is available.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.12
• /
• pp.1880-1889
• /
• 2022

The online portal platform provides online news with online comments, but the anonymity of comments causes incivility, and online comments are considered social problems. While there are many foreign language-based incivility detection studies, in-depth research is not being conducted in Korea since there has not been implemented Korean language dataset which is labeled detailed criteria of incivility. In this study, the incivility notation of comments was conducted in a total of 13 items, uncivil words were summarized. Furthermore, Attention algorithm was applied to each comment and summary to extract embedding vectors. 2-d CNN followed at the end to detect incivility in given data. As a result, we showed that the proposed algorithm is useful for anti-citizen detection such as name-calling and offensive tones. This study is expected to contribute to the formation of a healthy online comment culture by detecting uncivil comments which hinder democratic discourse.