• Title/Summary/Keyword: Fusion Model

Search Result 971, Processing Time 0.025 seconds

Welfare Interface using Multiple Facial Features Tracking (다중 얼굴 특징 추적을 이용한 복지형 인터페이스)

  • Ju, Jin-Sun;Shin, Yun-Hee;Kim, Eun-Yi
    • Journal of the Institute of Electronics Engineers of Korea SP
    • /
    • v.45 no.1
    • /
    • pp.75-83
    • /
    • 2008
  • We propose a welfare interface using multiple fecial features tracking, which can efficiently implement various mouse operations. The proposed system consist of five modules: face detection, eye detection, mouth detection, facial feature tracking, and mouse control. The facial region is first obtained using skin-color model and connected-component analysis(CCs). Thereafter the eye regions are localized using neutral network(NN)-based texture classifier that discriminates the facial region into eye class and non-eye class, and then mouth region is localized using edge detector. Once eye and mouth regions are localized they are continuously and correctly tracking by mean-shift algorithm and template matching, respectively. Based on the tracking results, mouse operations such as movement or click are implemented. To assess the validity of the proposed system, it was applied to the interface system for web browser and was tested on a group of 25 users. The results show that our system have the accuracy of 99% and process more than 21 frame/sec on PC for the $320{\times}240$ size input image, as such it can supply a user-friendly and convenient access to a computer in real-time operation.

Microstructural Characterization of Clad Interface in Welds of Ni-Cr-Mo High Strength Low Alloy Steel (Ni-Cr-Mo계 고강도 저합금강 용접클래드 계면의 미세조직 특성 평가)

  • Kim, Hong-Eun;Lee, Ki-Hyoung;Kim, Min-Chul;Lee, Ho-Jin;Kim, Keong-Ho;Lee, Chang-Hee
    • Korean Journal of Metals and Materials
    • /
    • v.49 no.8
    • /
    • pp.628-634
    • /
    • 2011
  • SA508 Gr.4N Ni-Cr-Mo low alloy steel, in which Ni and Cr contents are higher than in commercial SA508 Gr.3 Mn-Mo-Ni low alloy steels, may be a candidate reactor pressure vessel (RPV) material with higher strength and toughness from its tempered martensitic microstructure. The inner surface of the RPV is weld-cladded with stainless steels to prevent corrosion. The goal of this study is to evaluate the microstructural properties of the clad interface between Ni-Cr-Mo low alloy steel and stainless weldment, and the effects of post weld heat treatment (PWHT) on the properties. The properties of the clad interface were compared with those of commercial Mn-Mo-Ni low alloy steel. Multi-layer welding of model alloys with ER308L and ER309L stainless steel by the SAW method was performed, and then PWHT was conducted at $610^{\circ}C$ for 30 h. The microstructural changes of the clad interface were analyzed using OM, SEM and TEM, and micro-Vickers hardness tests were performed. Before PWHT, the heat affected zone (HAZ) showed higher hardness than base and weld metals due to formation of martensite after welding in both steels. In addition, the hardness of the HAZ in Ni-Cr-Mo low alloy steel was higher than that in Mn-Mo-Ni low alloy steel due to a comparatively high martensite fraction. The hardness of the HAZ decreased after PWHT in both steels, but the dark region was formed near the fusion line in which the hardness was locally high. In the case of Mn-Mo-Ni low alloy steel, formation of fine Cr-carbides in the weld region near the fusion line by diffusion of C from the base metal resulted in locally high hardness in the dark region. However, the precipitates of the region in the Ni-Cr-Mo low alloy steel were similar to that in the base metal, and the hardness in the region was not greatly different from that in the base metal.

A Study on Improving Facial Recognition Performance to Introduce a New Dog Registration Method (새로운 반려견 등록방식 도입을 위한 안면 인식 성능 개선 연구)

  • Lee, Dongsu;Park, Gooman
    • Journal of Broadcast Engineering
    • /
    • v.27 no.5
    • /
    • pp.794-807
    • /
    • 2022
  • Although registration of dogs is mandatory according to the revision of the Animal Protection Act, the registration rate is low due to the inconvenience of the current registration method. In this paper, a performance improvement study was conducted on the dog face recognition technology, which is being reviewed as a new registration method. Through deep learning learning, an embedding vector for facial recognition of a dog was created and a method for identifying each dog individual was experimented. We built a dog image dataset for deep learning learning and experimented with InceptionNet and ResNet-50 as backbone networks. It was learned by the triplet loss method, and the experiments were divided into face verification and face recognition. In the ResNet-50-based model, it was possible to obtain the best facial verification performance of 93.46%, and in the face recognition test, the highest performance of 91.44% was obtained in rank-5, respectively. The experimental methods and results presented in this paper can be used in various fields, such as checking whether a dog is registered or not, and checking an object at a dog access facility.

Detection of Cavities Behind Concrete Walls Using a Microphone (마이크로폰을 이용한 콘크리트 벽체 배면의 공동 탐사)

  • Kang, Seonghun;Lee, Jong-Sub;Han, WooJin;Kim, Sang Yeob;Yu, Jung-Doung
    • Journal of the Korean Geotechnical Society
    • /
    • v.38 no.12
    • /
    • pp.19-28
    • /
    • 2022
  • Cavities behind concrete walls can adversely affect the stability of structures. Thus study aims to detect cavities behind concrete structures using a microphone in a laboratory model test. A small-scale concrete wall is constructed in a chamber, which is composed of a reinforced concrete plate and dry soil. A plastic bowl is then placed between the plate and soil to simulate a cavity behind the concrete structure. Leaky surface acoustic waves are generated by impacting the concrete plate using a hammer and are measured using a microphone. The measured signals are analyzed using natural frequencies, and cavity-free sections are evaluated. The test results show that the first natural frequency decreases at the cavity section due to the flexural vibration behavior of the plate. In addition, the amplitude corresponding to the first natural frequency decreases as the measurement location becomes farther from the cavity center and significantly decreases at the measurement locations near the rebars. This study demonstrates that a microphone may be useful to detect cavities behind concrete walls.

Improvement of Silkworm Egg Microinjection Using 3D Printing Technology (3D 프린팅 기술을 이용한 누에 알 미세주입 기술 개선)

  • Jeong, Chan Young;Lee, Chang Hoon;Seok, Young-Seek;Yong, Sang Yeop;Kim, Seong-Wan;Kim, Kee Young;Park, Jong Woo
    • Korean journal of applied entomology
    • /
    • v.61 no.1
    • /
    • pp.249-254
    • /
    • 2022
  • Silkworms, which have for long been used as an insect resource for industrialization, have recently attracted attention as potential bio-factories for the production of novel biomaterials. In this regard, material production is typically achieved based on transformation technology, mediated via microinjection, in which a target gene is inserted into eggs containing an embryo. However, an essential step in the microinjection procedure is egg fixation, which can be a time-consuming and laborious task. Therefore, in this study, using the 3DCADian program, we adopted a 3D printing approach to model egg liners and glue drawers, which can contribute to facilitating egg alignment and fixation, thereby enhancing transformation efficiency by reducing time consumption and fatigue. After rendering using Fusion 360, the two supplementary tools were produced by printing with nylon resin (PA12) and Sinterit Lisa Pro. Subsequent analysis of the time required to fix eggs on glass slides using the two manufactured tools, revealed that the processing time was reduced by approximately 18.6% when the two tools were used compared with when these tools were not used. These innovations not only reduced fatigue but also contributed to more effective use of the microscope and manipulator for microinjection. Consequently, we believe that with additional research and refinement, the egg liner and glue drawer developed in this study could be used to enhance silkworm transformation efficiency and study similar transformation systems in other industrial insects.

Analysis of Transfer Learning Effect for Automatic Dog Breed Classification (반려견 자동 품종 분류를 위한 전이학습 효과 분석)

  • Lee, Dongsu;Park, Gooman
    • Journal of Broadcast Engineering
    • /
    • v.27 no.1
    • /
    • pp.133-145
    • /
    • 2022
  • Compared to the continuously increasing dog population and industry size in Korea, systematic analysis of related data and research on breed classification methods are very insufficient. In this paper, an automatic breed classification method is proposed using deep learning technology for 14 major dog breeds domestically raised. To do this, dog images are collected for deep learning training and a dataset is built, and a breed classification algorithm is created by performing transfer learning based on VGG-16 and Resnet-34 as backbone networks. In order to check the transfer learning effect of the two models on dog images, we compared the use of pre-trained weights and the experiment of updating the weights. When fine tuning was performed based on VGG-16 backbone network, in the final model, the accuracy of Top 1 was about 89% and that of Top 3 was about 94%, respectively. The domestic dog breed classification method and data construction proposed in this paper have the potential to be used for various application purposes, such as classification of abandoned and lost dog breeds in animal protection centers or utilization in pet-feed industry.

Enhanced Indoor Localization Scheme Based on Pedestrian Dead Reckoning and Kalman Filter Fusion with Smartphone Sensors (스마트폰 센서를 이용한 PDR과 칼만필터 기반 개선된 실내 위치 측위 기법)

  • Harun Jamil;Naeem Iqbal;Murad Ali Khan;Syed Shehryar Ali Naqvi;Do-Hyeun Kim
    • Journal of Internet of Things and Convergence
    • /
    • v.10 no.4
    • /
    • pp.101-108
    • /
    • 2024
  • Indoor localization is a critical component for numerous applications, ranging from navigation in large buildings to emergency response. This paper presents an enhanced Pedestrian Dead Reckoning (PDR) scheme using smartphone sensors, integrating neural network-aided motion recognition, Kalman filter-based error correction, and multi-sensor data fusion. The proposed system leverages data from the accelerometer, magnetometer, gyroscope, and barometer to accurately estimate a user's position and orientation. A neural network processes sensor data to classify motion modes and provide real-time adjustments to stride length and heading calculations. The Kalman filter further refines these estimates, reducing cumulative errors and drift. Experimental results, collected using a smartphone across various floors of University, demonstrate the scheme's ability to accurately track vertical movements and changes in heading direction. Comparative analyses show that the proposed CNN-LSTM model outperforms conventional CNN and Deep CNN models in angle prediction. Additionally, the integration of barometric pressure data enables precise floor level detection, enhancing the system's robustness in multi-story environments. Proposed comprehensive approach significantly improves the accuracy and reliability of indoor localization, making it viable for real-world applications.

Calculating the Sunlight Amount for Buildings Using SAS: A Case Study of Gyeongsan City (그림자 분석 시뮬레이션을 활용한 건축물별 일조량 산정 - 경산시를 사례로)

  • Kim, Do-Ryeong;Kim, Sung-Jae;Han, Soo-Hee;Jo, Myung-Hee
    • Journal of the Korean Association of Geographic Information Studies
    • /
    • v.17 no.1
    • /
    • pp.159-172
    • /
    • 2014
  • As greenhouse gas emissions have been increasing in the world, global warming is being recognized as a cause of the global problems like climate change. This is closely associated the fossil fuels. Thus renewable energy has been brought to the attention of many people as the upcoming alternative energy source to cope with the fossil drain and increased environmental regulations. Especially, the solar energy among renewable energy has drastically increased. In this study, we calculate on daylight ratio about the solar energy for buildings based on digital surface model. The digital surface model was made using the spatial information data. And it was simulated the shadow analysis using SAS. Therefore, it was suitable places to utilize the solar energy in the Gyeongsan city. Consequently, the daylight ratio was considered important factor to select region of the industry of the solar light power generation.

Laboratory Validation of Bridge Finite Model Updating Approach By Static Load Input/Deflection Output Measurements (정적하중입력/변위출력관계를 이용한 단경간 교량의 유한요소모델개선기법: 실내실험검증)

  • Kim, Sehoon;Koo, Ki Young;Lee, Jong-Jae
    • Journal of the Korea institute for structural maintenance and inspection
    • /
    • v.20 no.3
    • /
    • pp.10-17
    • /
    • 2016
  • This paper presents a laboratory validation of a new approach for Finite Element Model Updating(FEMU) on short-span bridges by combining ambient vibration measurements with static load input-deflection output measurements. The conventional FEMU approach based on modal parameters requires the assumption on the system mass matrix for the eigen-value analysis. The proposed approach doesn't require the assumption and even provides a way to update the mass matrix. The proposed approach consists of two steps: 1) updating the stiffness matrix using the static input-deflection output measurements, and 2) updating the mass matrix using a few lower natural frequencies. For a validation of the proposed approach, Young's modulus of the laboratory model was updated by the proposed approach and compared with the value obtained from strain-stress tests in a Universal Testing Machine. Result of the conventional FEMU was also compared with the result of the proposed approach. It was found that proposed approach successfully estimated the Young's modulus and the mass density reasonably while the conventional FEMU showed a large error when used with higher-modes. In addition, the FE modeling error was discussed.

Modeling Virtual Ecosystems that Consist of Artificial Organisms and Their Environment (인공생명체와 그들을 둘러싸는 환경으로 구성 되어지는 가상생태계 모델링)

  • Lee, Sang-Hee
    • Korean Journal of Agricultural and Forest Meteorology
    • /
    • v.12 no.2
    • /
    • pp.122-131
    • /
    • 2010
  • This paper introduces the concept of a virtual ecosystem and reports the following three mathematical approaches that could be widely used to construct such an ecosystem, along with examples: (1) a molecular dynamics simulation approach for animal flocking behavior, (2) a stochastic lattice model approach for termite colony behavior, and (3) a rule-based cellular automata approach for biofilm growth. The ecosystem considered in this study consists of artificial organisms and their environment. Each organism in the ecosystem is an agent that interacts autonomously with the dynamic environment, including the other organisms within it. The three types of model were successful to account for each corresponding ecosystem. In order to accurately mimic a natural ecosystem, a virtual ecosystem needs to take many ecological variables into account. However, doing so is likely to introduce excess complexity and nonlinearity in the analysis of the virtual ecosystem's dynamics. Nonetheless, the development of a virtual ecosystem is important, because it can provide possible explanations for various phenomena such as environmental disturbances and disasters, and can also give insights into ecological functions from an individual to a community level from a synthetic viewpoint. As an example of how lower and higher levels in an ecosystem can be connected, this paper also briefly discusses the application of the second model to the simulation of a termite ecosystem and the influence of climate change on the termite ecosystem.