• Title/Summary/Keyword: Feature-level

Search Result 1,273, Processing Time 0.031 seconds

Detection of Mammographic Microcalcifications by Statistical Pattern Classification 81 Pattern Matching (통계적 패턴 분류법과 패턴 매칭을 이용한 유방영상의 미세석회화 검출)

  • 양윤석;김덕원;김은경
    • Journal of Biomedical Engineering Research
    • /
    • v.18 no.4
    • /
    • pp.357-364
    • /
    • 1997
  • The early detection of breast cancer is clearly a key ingredient for reducing breast cancer mortality. Microcalcification is the only visible feature of the DCIS's(ductal carcinoma in situ) which consist 15 ~ 20% of screening-detected breast cancer. Therefore, the analysis of the shapes and distributions of microcalcifications is very significant for the early detection. The automatic detection procedures have b(:on the concern of digital image processing for many years. We proposed here one efficient method which is essentially statistical pattern classification accelerated by one representative feature, correlation coefficient. We compared the results by this additional feature with results by a simple gray level thresholding. The average detection rate was increased from 48% by gray level feature only to 83% by the proposed method The performances were evaluated with TP rates and FP counts, and also with Bayes errors.

  • PDF

Convolutional Neural Network Based Multi-feature Fusion for Non-rigid 3D Model Retrieval

  • Zeng, Hui;Liu, Yanrong;Li, Siqi;Che, JianYong;Wang, Xiuqing
    • Journal of Information Processing Systems
    • /
    • v.14 no.1
    • /
    • pp.176-190
    • /
    • 2018
  • This paper presents a novel convolutional neural network based multi-feature fusion learning method for non-rigid 3D model retrieval, which can investigate the useful discriminative information of the heat kernel signature (HKS) descriptor and the wave kernel signature (WKS) descriptor. At first, we compute the 2D shape distributions of the two kinds of descriptors to represent the 3D model and use them as the input to the networks. Then we construct two convolutional neural networks for the HKS distribution and the WKS distribution separately, and use the multi-feature fusion layer to connect them. The fusion layer not only can exploit more discriminative characteristics of the two descriptors, but also can complement the correlated information between the two kinds of descriptors. Furthermore, to further improve the performance of the description ability, the cross-connected layer is built to combine the low-level features with high-level features. Extensive experiments have validated the effectiveness of the designed multi-feature fusion learning method.

Speech emotion recognition based on genetic algorithm-decision tree fusion of deep and acoustic features

  • Sun, Linhui;Li, Qiu;Fu, Sheng;Li, Pingan
    • ETRI Journal
    • /
    • v.44 no.3
    • /
    • pp.462-475
    • /
    • 2022
  • Although researchers have proposed numerous techniques for speech emotion recognition, its performance remains unsatisfactory in many application scenarios. In this study, we propose a speech emotion recognition model based on a genetic algorithm (GA)-decision tree (DT) fusion of deep and acoustic features. To more comprehensively express speech emotional information, first, frame-level deep and acoustic features are extracted from a speech signal. Next, five kinds of statistic variables of these features are calculated to obtain utterance-level features. The Fisher feature selection criterion is employed to select high-performance features, removing redundant information. In the feature fusion stage, the GA is is used to adaptively search for the best feature fusion weight. Finally, using the fused feature, the proposed speech emotion recognition model based on a DT support vector machine model is realized. Experimental results on the Berlin speech emotion database and the Chinese emotion speech database indicate that the proposed model outperforms an average weight fusion method.

Geometric Model Decimation Method for Salient Features (돌출된 특징을 위한 기하 모델 단순화 방법)

  • Kim, Soo-Kyun;An, Sung-Og
    • The Journal of Korean Association of Computer Education
    • /
    • v.11 no.4
    • /
    • pp.85-93
    • /
    • 2008
  • This paper proposes a method for generating low-level geometric models with retaining salient features during decimation. Our method employs feature extraction technique for extracting feature lines defined via curvature derivatives on the model (we divide features into ridges and valleys). We add the extraction method to simplification technique (Feature Quadric Error Metric) for making coarse model with features. This paper clearly shows that experimental results have better quality and smaller geometric error than previous methods.

  • PDF

LxBSM: Loadable Kernel Module for the Creation of C2 Level Audit Data based on Linux (LxBSM: C2 수준의 감사 자료 생성을 위한 리눅스 기반 동적 커널 모듈)

  • 전상훈;최재영;김세환;심원태
    • Journal of KIISE:Computing Practices and Letters
    • /
    • v.10 no.2
    • /
    • pp.146-155
    • /
    • 2004
  • Currently most of commercial operating systems contain a high-level audit feature to increase their own security level. Linux does not fall behind the other commercial operating systems in performance and stability, but Linux does not have a good audit feature. Linux is required to support a higher security feature than C2 level of the TCSEC in order to be used as a server operating system, which requires the kernel-level audit feature that provides the system call auditing feature and audit event. In this paper, we present LxBSM, which is a kernel module to provide the kernel-level audit features. The audit record format of LxBSM is compatible with that of Sunshield BSM. The LxBSM is implemented as a loadable kernel module, so it has the enhanced usability. It provides the rich audit records including the user-level audit events such as login/logout. It supports both the pipe and file interface for increasing the connectivity between LxBSM and intrusion detection systems (IDS). The performance of LxBSM is compared and evaluated with that of Linux kernel without the audit features. The response time was increased when the system calls were called to create the audit data, such as fork, execve, open, and close. However any other performance degradation was not observed.

Assessment of Job stress and Psychosocial stress level using Psychosocial health measurement tool in dental technicians (사회심리적 건강측정도구를 이용한 치과기공사의 스트레스 평가)

  • Kim, Wook-Tae;Han, Tae-Young
    • Journal of Technologic Dentistry
    • /
    • v.31 no.3
    • /
    • pp.67-85
    • /
    • 2009
  • This study aims to provide the research for dental technician's stress prevention and management with basic materials by understanding dental technician's psychosocial stress level and examining relevant factors. The subject of this study is 255 dental technologists who work mainly in Seoul Gyeonggi district for a month of April of 2009 and I conducted cross-sectional study through self administered survey. The contents of survey include general feature, occupational feature, health behavior feature. I used Karasek's Job Content Questionnaire, JCQ and Psychosocial well-being index, PWI-SF as means of measurement. To compare the level of dental technician's psychosocial stress, I conducted t-test and ANOVA and I measured the factors that are related with psychosocial stress symptom with step by step multiple regressive analysis. According to the result of Cronbach's a value which is yielded to verify the reliability of means of measurement, the reliability of concept is sufficient. The detailed result of this study is as follows. 1. According to the result of analyzing the stress symptom in accordance with general feature and occupational feature, those dental technologists who are older and not married, graduate from junior college, have lower position, work at university hospital or general hospital show lower stress(p<0.05). There is no difference in the level of psychosocial stress with regard to duty related feature, period of service, daily average working hours, monthly average pay. 2. With regard to health behavior feature, those dental technologists who control weight better and have meal more regularly show lower stress(p<0.05). Those dental technicians who smoke, drink liquid and take a suitable sleep show low stress but the difference does not have significance statistically. 3. With regard to the factors of stress in the workplace, those dental technicians who have lower duty related requirement, have higher duty related control ability, have higher social support, have less instability of employment and have less workload and physical burden show lower stress(p<0.05). 4. According to the result of analyzing the factors that influence dental technologist's stress symptom, social support has the most enormous influence on stress symptom. Unstable employment, regular exercise, regular eating, daily average sleeping hours and technological capacity are also important in this order. According to the result of this study, those dental technicians who have higher social support, less instability of employment, do exercise more regularly, take enough sleep more soundly and have higher technological capacity show lower psychosocial stress symptom. Therefore, to adjust appropriately the dental technician's stress and properly maintain and improve the dental technician's mental health, effective management plan that enables dental technicians to maintain smooth human relationships for dental technicians should be sought. In addition, heath education and health management for dental technicians should be given more thoroughly so that they can establish desirable health behavior in daily life.

  • PDF

Thermal Imagery-based Object Detection Algorithm for Low-Light Level Nighttime Surveillance System (저조도 야간 감시 시스템을 위한 열영상 기반 객체 검출 알고리즘)

  • Chang, Jeong-Uk;Lin, Chi-Ho
    • The Journal of The Korea Institute of Intelligent Transport Systems
    • /
    • v.19 no.3
    • /
    • pp.129-136
    • /
    • 2020
  • In this paper, we propose a thermal imagery-based object detection algorithm for low-light level nighttime surveillance system. Many features selected by Haar-like feature selection algorithm and existing Adaboost algorithm are often vulnerable to noise and problems with similar or overlapping feature set for learning samples. It also removes noise from the feature set from the surveillance image of the low-light night environment, and implements it using the lightweight extended Haar feature and adaboost learning algorithm to enable fast and efficient real-time feature selection. Experiments use extended Haar feature points to recognize non-predictive objects with motion in nighttime low-light environments. The Adaboost learning algorithm with video frame 800*600 thermal image as input is implemented with CUDA 9.0 platform for simulation. As a result, the results of object detection confirmed that the success rate was about 90% or more, and the processing speed was about 30% faster than the computational results obtained through histogram equalization operations in general images.

Reliability improvement of nonlinear ultrasonic modulation based fatigue crack detection using feature-level data fusion

  • Lim, Hyung Jin;Kim, Yongtak;Sohn, Hoon;Jeon, Ikgeun;Liu, Peipei
    • Smart Structures and Systems
    • /
    • v.20 no.6
    • /
    • pp.683-696
    • /
    • 2017
  • In this study, the reliability of nonlinear ultrasonic modulation based fatigue crack detection is improved using a feature-level data fusion approach. When two ultrasonic inputs at two distinct frequencies are applied to a specimen with a fatigue crack, modulation components at the summation and difference of these two input frequencies appear. First, the spectral amplitudes of the modulation components and their spectral correlations are defined as individual features. Then, a 2D feature space is constructed by combining these two features, and the presence of a fatigue crack is identified in the feature space. The effectiveness of the proposed fatigue crack detection technique is experimentally validated through cyclic loading tests of aluminum plates, full-scale steel girders and a rotating shaft component. Subsequently, the improved reliability of the proposed technique is quantitatively investigated using receiver operating characteristic analysis. The uniqueness of this study lies in (1) improvement of nonlinear ultrasonic modulation based fatigue crack detection reliability using feature-level data fusion, (2) reference-free fatigue crack diagnosis without using the baseline data obtained from the intact condition of the structure, (3) application to full-scale steel girders and shaft component, and (4) quantitative investigation of the improved reliability using receiver operating characteristic analysis.

Efficient Recognition Method for Ballistic Warheads by the Fusion of Feature Vectors Based on Flight Phase (비행 단계별 특성벡터 융합을 통한 효과적인 탄두 식별방법)

  • Choi, In-Oh;Kim, Si-Ho;Jung, Joo-Ho;Kim, Kyung-Tae;Park, Sang-Hong
    • The Journal of Korean Institute of Electromagnetic Engineering and Science
    • /
    • v.30 no.6
    • /
    • pp.487-497
    • /
    • 2019
  • It is very difficult to detect ballistic missiles because of small cross-sections of the radar and the high maneuverability of the missiles. In addition, it is very difficult to recognize and intercept warheads because of the existence of debris and decoy with similar motion parameters in each flight phase. Therefore, feature vectors based on the maneuver, the micro-motion according to flight phase are needed, and the two types of features must be fused for the efficient recognition of ballistic warhead regardless of the flight phase. In this paper, we introduce feature vectors appropriate for each flight phase and an effective method to fuse them at the feature vector-level and classifier-level. According to the classification simulations using the radar signals predicted by the CAD models, the closer the warhead was to the final destination, the more improved was the classification performance. This was achieved by the classifier-level fusion, regardless of the flight phase in a noisy environment.

An Input Feature Selection Method Applied to Fuzzy Neural Networks for Signal Estimation

  • Na, Man-Gyun;Sim, Young-Rok
    • Nuclear Engineering and Technology
    • /
    • v.33 no.5
    • /
    • pp.457-467
    • /
    • 2001
  • It is well known that the performance of a fuzzy neural network strongly depends on the input features selected for its training. In its applications to sensor signal estimation, there are a large number of input variables related with an output As the number of input variables increases, the training time of fuzzy neural networks required increases exponentially. Thus, it is essential to reduce the number of inputs to a fuzzy neural network and to select the optimum number of mutually independent inputs that are able to clearly define the input-output mapping. In this work, principal component analysis (PCA), genetic algorithms (CA) and probability theory are combined to select new important input features. A proposed feature selection method is applied to the signal estimation of the steam generator water level, the hot-leg flowrate, the pressurizer water level and the pressurizer pressure sensors in pressurized water reactors and compared with other input feature selection methods.

  • PDF