• Journal of the Korea Convergence Society
• /
• v.12 no.12
• /
• pp.31-37
• /
• 2021

In this paper proposed a feature fusion and spatial attention-based modified YOLOv4S for small and occluded detection. Conventional YOLOv4S is a lightweight network and lacks feature extraction capability compared to the method of the deep network. The proposed method first combines feature maps of different scales with feature fusion to enhance semantic and low-level information. In addition expanding the receptive field with dilated convolution, the detection accuracy for small and occluded objects was improved. Second by improving the conventional spatial information with spatial attention, the detection accuracy of objects classified and occluded between objects was improved. PASCAL VOC and COCO datasets were used for quantitative evaluation of the proposed method. The proposed method improved mAP by 2.7% in the PASCAL VOC dataset and 1.8% in the COCO dataset compared to the Conventional YOLOv4S.

• Journal of the Korean Society of Costume
• /
• v.58 no.2
• /
• pp.134-148
• /
• 2008

Methods of delivering products to consumers do not act as less important factor than creative activities to create new product designs. Mobilizing various psychological elements based on human instinct and desire, fashion industry presents a product as an object of envy, gaining viewers' attention. Here, the viewer does not simply take the product as an object to view, but also imagine transformation it will bring. The study of the cause and effects of the interaction, which is found in the relationship between "the viewer" and "what is viewed" is an important factor that needs to be identified in the phase of creation as well as in the aspect of delivery. The relationship between the perceiver and what is perceived features in designs, product advertisements, related articles, and fashion shows in modern fashion, serving as a medium that enables the humans, who must inevitably exist between the two poles such as mind and body, the subject and the object, the ego and non-ego, and the reality and an ideal, to communicate between the poles. Humans cannot do arbitrary acts or make arbitrary selections only as they access to foreign things through instinct, desire, or experienced perception, and they are sometimes positioned passive by things. In the background, as there are human dual characteristics in which they are expressed as the ego and another ego who exists inside of the ego, they not only view an object, but also become an object to be viewed. Many products in modern fashion, as the objects of reciprocal transposition, grow giving illusions to humans. Having a desire for such objects is human's instinct and normal act to keep the life balanced between the reality and an ideal, which is based on the activation of reality function. Furthermore, freely acting rather than ignoring or overcoming the desire may be the act of retrieving one's ownership to the ego.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.7
• /
• pp.2131-2153
• /
• 2022

To achieve accurate detection of tuberculosis (TB) areas in chest radiographs, we design a chest X-ray TB area detection algorithm. The algorithm consists of two stages: the chest X-ray TB classification network (CXTCNet) and the chest X-ray TB area detection network (CXTDNet). CXTCNet is used to judge the presence or absence of TB areas in chest X-ray images, thereby excluding the influence of other lung diseases on the detection of TB areas. It can reduce false positives in the detection network and improve the accuracy of detection results. In CXTCNet, we propose a channel attention mechanism (CAM) module and combine it with DenseNet. This module enables the network to learn more spatial and channel features information about chest X-ray images, thereby improving network performance. CXTDNet is a design based on a sparse object detection algorithm (Sparse R-CNN). A group of fixed learnable proposal boxes and learnable proposal features are using for classification and location. The predictions of the algorithm are output directly without non-maximal suppression post-processing. Furthermore, we use CLAHE to reduce image noise and improve image quality for data preprocessing. Experiments on dataset TBX11K show that the accuracy of the proposed CXTCNet is up to 99.10%, which is better than most current TB classification algorithms. Finally, our proposed chest X-ray TB detection algorithm could achieve AP of 45.35% and AP50 of 74.20%. We also establish a chest X-ray TB dataset with 304 sheets. And experiments on this dataset showed that the accuracy of the diagnosis was comparable to that of radiologists. We hope that our proposed algorithm and established dataset will advance the field of TB detection.

• Journal of Korean Library and Information Science Society
• /
• v.24
• /
• pp.107-161
• /
• 1996

This study aims at introducing 'ethnography' which is one of the most representative qualitative research methods into Library and Information Science research in Korea.. Ethnography, derived from anthropology, helps researchers to understand the whole and deep aspect of the research object. That is because the researcher puts himself into the life-world of the research object and observes it for a long time. Ethnography can be used as an alternative method to quantitative research methods. Until now, Library and Information Science research in Korea has used quantitative research methods in almost every case. From the 1980s so-called 'scientific methods' using hypotheses testing, have provided the major premise for research methodology in Library and Information Science of Korea. And the researchers have seen their research object(Korean Libraries) not in the native perspective but largely in the western(especially American) perspective. There is a need in Korea for more culturally relative research. So the desirability of introducing ethnography and other qualitative research methods into Library and Information Science research in Korea can be summarized as follows : I. Ethnography and other qualitative methods are needed for the researchers to overcome the limitation of quantitative methods which have formed the main methodological paradigm in Library and Information Science research in Korea. While those quantitative scientific methods can be a n.0, pplied to the social sciences, they are not adequate for the social sciences. It is because the research objects of the social sciences are human and social phenomena. II. It is needed that Library and Information Science research in Korea pay more attention to the speciality of Korean libraries. To do researches based on the viewpoint of cultural-relativism, researchers should consider the cultural context of Korean libraries. During the past years researchers in other social science fields in Korea, especially sociology and pedagogy, have gradually a n.0, pplied the methods of ethnography to their fields. These social scientists have attempted to escape from ethnocentrism, a problem which has greatly influenced past and present research methods. To get a holistic and in-depth understanding of Korean libraries on the present stage, and to solve their problems radically, it seems imperative that Library and Information Science research in Korea pay more attention to qualitative research methods such as ethnography.

• Korean Journal of Computational Design and Engineering
• /
• v.17 no.6
• /
• pp.417-425
• /
• 2012

In recent years, great attention has been paid to using simple physical objects as tangible objects to improve user interaction in augmented reality (AR) environments. In this paper, we address AR-based user interaction using tangible objects, which has been used as a key component for virtual design evaluation of engineered products including digital handheld products. We herein consider the use of two types (product-type and pointer-type) of tangible objects. The user creates input events by touching specified parts of the product-type object with the pointer-type object, and the virtual product reacts to the events by rendering its visual and auditory contents on the output devices. The product-type object is used to reflect the geometric shape of a product of interest and to determine its position and orientation in the AR environment. The pointer-type object is used to recognize the reference position of the pointer (or finger) in the same environment. The rapid prototype of the product is employed as a good alternative to the product-type object, but various alternatives to the pointer-type object can be considered according to fabrication process and touching mechanism. In this paper, we present four alternatives to the pointer-type object and investigate their strong and weak points by performing experimental comparison of their various aspects including interaction accuracy, task performance, and qualitative user experience.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.12
• /
• pp.561-568
• /
• 2021

In recent years, autonomous driving technologies have become a high-value-added technology that attracts attention in the fields of science and industry. For smooth Self-driving, it is necessary to accurately detect an object and estimate its movement speed in real time. CNN-based deep learning algorithms and conventional dense optical flows have a large consumption time, making it difficult to detect objects and estimate its movement speed in real time. In this paper, using a single camera image, fast object detection was performed using the YOLOv5 algorithm, a deep learning algorithm, and fast estimation of the speed of the object was performed by using a local dense optical flow modified from the existing dense optical flow based on the detected object. Based on this algorithm, we present a system that can predict the collision time and probability, and through this system, we intend to contribute to prevent tram accidents.

• 한국정보디스플레이학회:학술대회논문집
• /
• 2009.10a
• /
• pp.1022-1024
• /
• 2009

Stereoscopic Images provide depth information with the relative distances between the objects in the images. There are many different ways to extract disparity maps from the visible spectral images. For the infrared spectral range, the same approach cannot be utilized for the innate low resolution and colorless features because typical methods require corresponding features between the images. The authors suggest a new approach that makes use of image segmentation to obtain depth information for stereoscopic millimeter-wave images. For image segmentation a selective visual attention model based on the theory of a feature-integration of attention is used. Experimental results show the proposed method provides reasonable depth information for object shape recognition and display.

• Journal of Broadcast Engineering
• /
• v.27 no.3
• /
• pp.391-401
• /
• 2022

As deep learning-based object detection and recognition research have been developed recently, the scope of application to industry and real life is expanding. But deep learning-based systems in the construction system are still much less studied. Calculating materials in the construction system is still manual, so it is a reality that transactions of wrong volumn calculation are generated due to a lot of time required and difficulty in accurate accumulation. A fast and accurate automatic drawing recognition system is required to solve this problem. Therefore, we propose an AI-based automatic drawing recognition accumulation system that detects and recognizes steel materials in construction drawings. To accurately detect steel materials in construction drawings, we propose data augmentation techniques and spatial attention modules for improving small object detection performance based on YOLOv4. The detected steel material area is recognized by text, and the number of steel materials is integrated based on the predicted characters. Experimental results show that the proposed method increases the accuracy and precision by 1.8% and 16%, respectively, compared with the conventional YOLOv4. As for the proposed method, Precision performance was 0.938. The recall was 1. Average Precision AP_0.5 was 99.4% and AP_0.5:0.95 was 67%. Accuracy for character recognition obtained 99.9.% by configuring and learning a suitable dataset that contains fonts used in construction drawings compared to the 75.6% using the existing dataset. The average time required per image was 0.013 seconds in the detection, 0.65 seconds in character recognition, and 0.16 seconds in the accumulation, resulting in 0.84 seconds.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• fall
• /
• pp.113-116
• /
• 2021

일반적으로 딥러닝 기반의 객체 검출(Object Detection)기법은 합성곱 신경망(Convolutional Neural Network, CNN)을 통해 입력된 영상의 특징(Feature)을 추출하여 이를 통해 객체 검출을 수행한다. 최근 자연어 처리 분야에서 획기적인 성능을 보인 트랜스포머(Transformer)가 영상 분류, 객체 검출과 같은 컴퓨터 비전 작업을 수행하는데 있어 경쟁력이 있음이 드러나고 있다. 본 논문에서는 YOLOv4-CSP의 CSP 블록을 개선한 one-stage 방식의 객체 검출 네트워크를 제안한다. 개선된 CSP 블록은 트랜스포머(Transformer)의 멀티 헤드 어텐션(Multi-Head Attention)과 CSP 형태의 공간 피라미드 풀링(Spatial Pyramid Pooling, SPP) 연산을 기반으로 네트워크의 Backbone과 Neck에서의 feature 학습을 돕는다. 본 실험은 MSCOCO test-dev2017 데이터 셋으로 평가하였으며 제안하는 네트워크는 YOLOv4-CSP의 경량화 모델인 YOLOv4s-mish에 대하여 평균 정밀도(Average Precision, AP)기준 2.7% 향상된 검출 정확도를 보인다.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.21 no.3
• /
• pp.81-89
• /
• 2021

Under natural conditions, Apple detection has the problems of occlusion and small object detection difficulties. This paper proposes an improved model based on SSD. The SSD backbone network VGG16 is replaced with the ResNet50 network model, and the receptive field structure RFB structure is introduced. The RFB model amplifies the feature information of small objects and improves the detection accuracy of small objects. Combined with the attention mechanism (SE) to filter out the information that needs to be retained, the semantic information of the detection objectis enhanced. An improved SSD algorithm is trained on the VOC2007 data set. Compared with SSD, the improved algorithm has increased the accuracy of occlusion and small object detection by 3.4% and 3.9%. The algorithm has improved the false detection rate and missed detection rate. The improved algorithm proposed in this paper has higher efficiency.