• Asian Pacific Journal of Cancer Prevention
• /
• 제16권9호
• /
• pp.3867-3870
• /
• 2015

Background: To explore whether combined detection of serum tumor markers (CEA, CA72-4, CA19-9 and TSGF) improve the sensitivity and accuracy in the diagnosis of gastric cancer (GC). Materials and Methods: An automatic chemiluminescence immune analyzer with matched kits were used to determine the levels of serum CEA, CA72-4, CA19-9 and TSGF in 45 patients with gastric cancer (GC group), 40 patients with gastric benign diseases (GBD group) hospitalized in the same period and 30 healthy people undergoing a physical examination. The values of those 4 tumor markers in the diagnosis of gastric cancer was analyzed. Results: The levels of serum CEA, CA72-4, CA19-9 and TSGF of the GC group were higher than those of the GBD group and healthy examined people and the differences were significant (P<0.001). The area under receiver operating characteristic (ROC) curves for single detection of CEA, CA72-4, CA19-9 and TSGF in the diagnosis of GC was 0.833, 0.805, 0.810 and 0.839, respectively. The optimal cutoff values for these 4 indices were 2.36 ng/mL, 3.06 U/mL, 5.72 U/mL and 60.7 U/mL, respectively. With combined detection of tumor markers, the diagnostic power of those 4 indices was best, with an area under the ROC curve of 0.913 (95%CI 0.866~0.985), a sensitivity of 88.9% and a diagnostic accuracy of 90.4%. Conclusions: Combined detection of serum CEA, CA72-4, CA19-9 and TSGF increases the sensitivity and accuracy in diagnosis of GC, so it can be regarded as the important means for early diagnosis.

• 대한조선학회논문집
• /
• 제58권5호
• /
• pp.303-313
• /
• 2021

Recently, quality control of the Liquefied Natural Gas Carrier (LNGC) cargo hold and block-erection interference areas using 3D scanners have been performed, focusing on large shipyards and the international association of classification societies. In this study, as a part of the research on LNGC cargo hold quality management advancement, a study on deep-learning-based scaffolding system 3D point cloud object detection and post-processing were conducted using a LNGC cargo hold 3D point cloud. The scaffolding system point cloud object detection is based on the PointNet deep learning architecture that detects objects using point clouds, achieving 70% prediction accuracy. In addition, the possibility of improving the accuracy of object detection through parameter adjustment is confirmed, and the standard of Intersection over Union (IoU), an index for determining whether the object is the same, is achieved. To avoid the manual post-processing work, the object detection architecture allows automatic task performance and can achieve stable prediction accuracy through supplementation and improvement of learning data. In the future, an improved study will be conducted on not only the flat surface of the LNGC cargo hold but also complex systems such as curved surfaces, and the results are expected to be applicable in process progress automation rate monitoring and ship quality control.

• International Journal of Computer Science & Network Security
• /
• 제23권8호
• /
• pp.177-189
• /
• 2023

Malware detection is an increasingly important operational focus in cyber security, particularly given the fast pace of such threats (e.g., new malware variants introduced every day). There has been great interest in exploring the use of machine learning techniques in automating and enhancing the effectiveness of malware detection and analysis. In this paper, we present a deep recurrent neural network solution as a stacked Long Short-Term Memory (LSTM) with a pre-training as a regularization method to avoid random network initialization. In our proposal, we use global and short dependencies of the inputs. With pre-training, we avoid random initialization and are able to improve the accuracy and robustness of malware threat hunting. The proposed method speeds up the convergence (in comparison to stacked LSTM) by reducing the length of malware OpCode or bytecode sequences. Hence, the complexity of our final method is reduced. This leads to better accuracy, higher Mattews Correlation Coefficients (MCC), and Area Under the Curve (AUC) in comparison to a standard LSTM with similar detection time. Our proposed method can be applied in real-time malware threat hunting, particularly for safety critical systems such as eHealth or Internet of Military of Things where poor convergence of the model could lead to catastrophic consequences. We evaluate the effectiveness of our proposed method on Windows, Ransomware, Internet of Things (IoT), and Android malware datasets using both static and dynamic analysis. For the IoT malware detection, we also present a comparative summary of the performance on an IoT-specific dataset of our proposed method and the standard stacked LSTM method. More specifically, of our proposed method achieves an accuracy of 99.1% in detecting IoT malware samples, with AUC of 0.985, and MCC of 0.95; thus, outperforming standard LSTM based methods in these key metrics.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2002년도 ITC-CSCC -1
• /
• pp.46-49
• /
• 2002

One problem of vision-based eye-gazed detection is that it gives low resolution. Base on the displacement of the eyes, we proposed method for vision-based eye-gaze detection. While looking at difference positions on the screen, the distance of the centers of the eyes change accordingly. This relationship is derived and used to map the displacement to the distance in the screen. The experiments are performed to measure the accuracy and resolution to verify the proposed method. The results shown the accuracy of the screen mapping function for the horizontal plane are 76.47% and the error of this function be 23.53%

• 말소리와 음성과학
• /
• 제1권2호
• /
• pp.15-22
• /
• 2009

Prominence of speech, which is often called 'accent,' affects the fluency of speaking American English greatly. In this paper, we present an accurate prominence detection method that can be utilized in computer-aided language learning (CALL) systems. We employed pitch movement, overall syllable energy, 300-2200 Hz band energy, syllable duration, and spectral and temporal correlation as features to model the prominence of speech. After the features for vowel syllables of speech were extracted, prominent syllables were classified by SVM (Support Vector Machine). To further improve accuracy, the differences in characteristics of neighboring syllables were added as additional features. We also applied a speech recognizer to extract more precise syllable boundaries. The performance of our prominence detector was measured based on the Intonational Variation in English (IViE) speech corpus. We obtained 84.9% accuracy which is about 10% higher than previous research.

• 동력기계공학회지
• /
• 제19권2호
• /
• pp.33-39
• /
• 2015

In conducting indoor positioning by code division multiple access using spread spectrum ultrasonic waves, it is required to detect signals under the influence of near-far problem occurred by difference on signal power, caused by distance between transmitter and receiver. For discussing robustness to the problem, we verified measuring accuracy on distance from an experiment on a real space with a hardware device where our proposed method is mounted. The proposed method performs automatic signal detection by setting threshold level dynamically. As an experimental result, measurable distance were improved by the proposed method, and measurement errors were up to 50mm in distances from 1000mm to 6000mm; therefore, enough accuracy to realize self-localization or navigation for autonomous mobile robot or human was obtained.

• Journal of Multimedia Information System
• /
• 제4권1호
• /
• pp.43-50
• /
• 2017

In this paper, we propose a method which detects the nose and face of certain human by using the depth image. The proposed method has advantages of the low computational complexity and the high accuracy even in dark environment. Also, the detection accuracy of nose and face does not change in various postures. The proposed method first locates the locally protruding part from the depth image of the human body captured through the depth camera, and then confirms the nose through the depth characteristic of the nose and surrounding pixels. After finding the correct pixel of the nose, we determine the region of interest centered on the nose. In this case, the size of the region of interest is variable depending on the depth value of the nose. Then, face region can be found by performing binarization using the depth histogram in the region of interest. The proposed method can detect the nose and the face accurately regardless of the pose or the illumination of the captured area.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2006년도 제37회 하계학술대회 논문집 D
• /
• pp.1771-1772
• /
• 2006

In this paper, we considered multiple faults detection on a coaxial cable through Time-Frequency Domain Reflectometry (TFDR). It is well known that TFDR has high resolution accuracy for detecting and estimating the fault detection on a coaxial cable. This approach was based on time-frequency signal analysis and utilized a chirp signal multiplied by a Gaussian time envelope. The Gaussian envelope provided time localization, while the chirp allowed one to excite the system interest. We carried out experiments with 10C-FBT coaxial cable having either one or two faults. The result shows TFDR can be extended to detect multiple faults with high accuracy on a coaxial cable.

• JSTS:Journal of Semiconductor Technology and Science
• /
• 제17권1호
• /
• pp.79-85
• /
• 2017

In this paper, several optimizations are proposed to enhance the quality of lane detection algorithms in automotive applications. Considering the diagonal directions of lanes, the proposed limited Hough transform newly introduces image-splitting and angle-limiting schemes that relax the number of possible angles at the line voting process. In addition, unnecessary edges along the horizontal and vertical directions are pre-defined and removed during the edge detection procedures, increasing the detecting accuracy remarkably. Simulation results shows that the proposed lane recognition algorithm achieves an accuracy of more than 90% and a computing speed of 92 frame/sec, which are superior to the results from the previous algorithms.

• 전자공학회논문지S
• /
• 제34S권11호
• /
• pp.126-137
• /
• 1997

This paper describes the algorithm which can estimate road following direction and deetect obstacle using a monocular vision system. This algorithm can estimate the course of vehicle using the vanishing point properties and detect obstacle by statistical method. The proposed algorithm is composed of four steps, which are lane prediction, lane extraction, road following parameter estimation and obstacle detection. It is designed for high processing speed and high accuracy. The former is achieved by a small area named sub-windown in lane existence area, the later is realized by using connected edge points of lane. We would like to present that the new mehod can detect obstacle using the simple statistical method. The paracticalities of the processing speed, the accuracy of the algorithm and proposing obstacle detection method, have been justified through the experiment applied VTR image of the real road to the algorithm.