• The Journal of the Korean bone and joint tumor society
• /
• v.11 no.1
• /
• pp.32-39
• /
• 2005

Introduction: Currently, F-18 fluorodeoxyglucose positron emission tomography scans (FDG-PET) has been investigated in soft tissue tumor especially for tumor detection and noninvasive grading. However, the validity and the efficacy of FDG-PET are still unclear in clinical evaluation. The purpose of this study is to determine the efficacy of FDG-PET in compared to conventional diagnostic imaging studies currently used in the soft tissue tumor. Methods: Between March 2001 and March 2002, 29 patients (sixteen males, thirteen females, mean age, 47 years; a range from 4 to 73) diagnosed with soft tissue tumor were evaluated by both conventional diagnostic imaging and FDG-PET. Valid reference test of the local lesion was the histopathologic diagnosis, which was measured in all patients. The suspecting metastasis in the imaging studies was validated by pathology or follow up imaging for at least 6 months. Each imaging diagnosis was made independently. The accuracy of each diagnostic method was evaluated. The incremental cost accuracy ratio was determined in each diagnostic method. Results: For detection of local lesion, sensitivity, specificity, and accuracy for MRI and FDGPET scans were 91%, 57%, 83% and 95%, 43%, 83% respectively. For detection of distant lesion, sensitivity, specificity, accuracy for conventional diagnostic methods and FDG-PET scans were 77%, 89%, 87% and 92%, 94%, 93% respectively. The incremental cost accuracy ratio (ICAR) of FDG-PET for detection of distant lesion was 145,000won/%. According to ICAR for each tumor grade, PET strategy is most cost-effective at high grade tumors. Conclusions: For detection of local lesion such as recurrence or remnant tumor, FDG-PET scan was not more accurate than MRI. However, It was more accurate for detection of metastatic lesion than conventional methods. For detection of high grade tumor, PET was most costeffective than for detection of lower grade tumor.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.3
• /
• pp.19-26
• /
• 2020

Steel production requires high-quality surfaces with minimal defects. Therefore, the detection algorithms for the surface defects of steel strip should have good generalization performance. To meet the growing demand for high-quality products, the use of intelligent visual inspection systems is becoming essential in production lines. In this paper, we proposed a ShuffleDefectNet defect detection system based on deep learning. The proposed defect detection system exceeds state-of-the-art performance for defect detection on the Northeastern University (NEU) dataset obtaining a mean average accuracy of 99.75%. We train the best performing detection with different amounts of training data and observe the performance of detection. We notice that accuracy and speed improve significantly when use the overall architecture of ShuffleDefectNet.

• Smart Media Journal
• /
• v.4 no.4
• /
• pp.9-17
• /
• 2015

Recently the need for advanced security technologies are increasing as the occurrence of intelligent crime is growing fastly. Previous liveness detection and fake face detection methods are required for the improvement of accuracy in order to be put to practical use. In this paper, we propose a new liveness detection method using pupil reflection, and new fake image detection using Adaboost detector. The proposed system detects eyes based on multi-scale Gabor feature vector in the first stage, The template matching plays a role in determining the allowed eye area. And then, the reflected image in the pupil is used to decide whether or not the captured image is live or not. Experimental results indicate that the proposed method is superior to the previous methods in the detection accuracy of fake images.

• Journal of Korea Multimedia Society
• /
• v.19 no.8
• /
• pp.1310-1319
• /
• 2016

Face detection is the first step in a wide range of face applications. However, detecting faces in the wild is still a challenging task due to the wide range of variations in pose, scale, and occlusions. Recently, many deep learning methods have been proposed for face detection. However, further improvements are required in the wild. Another important issue to be considered in the face detection is the computational complexity. Current state-of-the-art deep learning methods require a large number of patches to deal with varying scales and the arbitrary image sizes, which result in an increased computational complexity. To reduce the complexity while achieving better detection accuracy, we propose a fully convolutional network-based face detection that can take arbitrarily-sized input and produce feature maps (heat maps) corresponding to the input image size. To deal with the various face scales, a multi-scale network architecture that utilizes the facial components when learning the feature maps is proposed. On top of it, we design multi-task learning technique to improve detection performance. Extensive experiments have been conducted on the FDDB dataset. The experimental results show that the proposed method outperforms state-of-the-art methods with the accuracy of 82.33% at 517 false alarms, while improving computational efficiency significantly.

• Journal of Korean Society of Occupational and Environmental Hygiene
• /
• v.7 no.2
• /
• pp.223-232
• /
• 1997

The accuracy and precision of a modified method of NIOSH Method 7600 and EPA method 218.6 was determined for analyzing hexavalent chromium, Cr(VI), collected on PVC filter from workplace air. The method was designed to extract from Cr(VI) on PVC filter with a alkali solution, 2% NaOH/3% $Na_2CO_3$, and to analyze it using ion chromatography/visible absorbance detection(IC/VAD). The results and conclusion are as the following. 1. The peak of Cr(VI) was separated sharply on chromatogram and was linearly related with Cr(VI) concentration in sloution. The correlation coefficient was 0.9999 in a calibration curve. The limit of detection was 0.25 $0.25{\mu}g/sample$. 2. The accuracy(% recovery) was 93.3% in a set of sample($9-50{\mu}g$) stored for a day, and 100.1%($10-60{\mu}g$) in another set of samples stored for 2 hours. It is assumed that the difference in recovery by storage time was due to reduction of Cr(VI) to Cr(III). 3. The precision(coefficient of variation, CV) of the method was 0.015 in spiked samples with Cr(VI) standard solution, and 0.010 in spiked samples with plating solution from a chrome electroplating factory. The overall CV in all types of samples was 0.0013. 4. The Cr(VI) was stable in 2% NaOH/3% $Na_2CO_3$ at least for 10 hours. In conclusion, the IC/VAD method is appropriate for determining low-level Cr(VI) in workplace air containing various interferences.

• Asian Pacific Journal of Cancer Prevention
• /
• v.15 no.18
• /
• pp.7981-7984
• /
• 2014

Purpose: Colorectal cancer (CRC) screening with fecal occult blood testing (FOBT) has been associated with a reduction in CRC incidence and CRC-related mortality. However, a conventional FOBT requires stool collection and handling, which may be inconvenient for participants. The EZ-Detect$^{TM}$ (Siam Pharmaceutical Thailand) is a FDA-approved chromogen-substrate based FOBT which is basically a self-checked FOBT (no stool handling required). This study aimed to evaluate the accuracy of EZ-Detect for CRC detection. Methods: This prospective study was conducted in the Faculty of Medicine, Siriraj Hospital, Bangkok, Thailand between November 2013 and May 2014. Some 96 patients with histologically-proven CRC and 101 patients with normal colonoscopic findings were invited to perform self-checked FOBT according to the manufacturer's instructions. Results were compared with endoscopic and pathologic findings. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for CRC detection were calculated. Results: The present study revealed the sensitivity, specificity, PPV and NPV of this self-checked FOBT for CRC detection to be 41% (95% CI: 31-51), 97% (95% CI: 92-99), 93% (95% CI: 81-98) and 63% (95% CI: 55-70), respectively. The overall accuracy of the self-checked FOBT for identifying CRC was 70%. The sensitivity for CRC detection based on 7th AJCC staging was 29% for stage I, 32% for stage II and 50% for stage III/IV (P=0.19). The sensitivity was 33% for proximal colon and 42% for distal colon and rectal cancer (P=0.76). Notably, none of nine infiltrative lesions gave a positive FOBT. Conclusions: The self-checked FOBT had an acceptable accuracy of CRC detection except for infiltrative tumors. This home-administrated or 'DIY' do-it-yourself FOBT could be considered as one non-invasive and convenient tool for CRC screening.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.11
• /
• pp.2727-2732
• /
• 2014

Recently, the researches of object tracking and recognition using Microsoft's Kinect are being actively studied. In this environment human hand detection and tracking is the most basic technique for human computer interaction. This paper proposes a method of improving the accuracy of the detected hand region's boundary in the cluttered background. To do this, we combine the hand detection results using the skin color with the extracted depth image from Kinect. From the experimental results, we show that the proposed method increase the accuracy of the hand region detection than the method of detecting a hand region with a depth image only. If the proposed method is applied to the sign language or gesture recognition system it is expected to contribute much to accuracy improvement.

• KIPS Transactions on Computer and Communication Systems
• /
• v.11 no.11
• /
• pp.411-418
• /
• 2022

Incorrectly loaded containers can easily knock down by strong winds. Container collapse accidents can lead to material damage and paralysis of the port system. In this paper, We propose a deep learning-based container loading state and accident risk detection technique. Using Darknet-based YOLO, the container load status identifies in real-time through corner casting on the top and bottom of the container, and the risk of accidents notifies the manager. We present criteria for classifying container alignment states and select efficient learning algorithms based on inference speed, classification accuracy, detection accuracy, and FPS in real embedded devices in the same environment. The study found that YOLOv4 had a weaker inference speed and performance of FPS than YOLOv3, but showed strong performance in classification accuracy and detection accuracy.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.17 no.4
• /
• pp.577-586
• /
• 2022

Detection and classification of steel surface defects are critical for product quality control in the steel industry. However, due to its low accuracy and slow speed, the traditional approach cannot be effectively used in a production line. The current, widely used algorithm (based on deep learning) has an accuracy problem, and there are still rooms for development. This paper proposes a method of steel surface defect detection combining EfficientNetV2 for image classification and YOLOv5 as an object detector. Shorter training time and high accuracy are advantages of this model. Firstly, the image input into EfficientNetV2 model classifies defect classes and predicts probability of having defects. If the probability of having a defect is less than 0.25, the algorithm directly recognizes that the sample has no defects. Otherwise, the samples are further input into YOLOv5 to accomplish the defect detection process on the metal surface. Experiments show that proposed model has good performance on the NEU dataset with an accuracy of 98.3%. Simultaneously, the average training speed is shorter than other models.

• Journal of Korean Society of Forest Science
• /
• v.81 no.4
• /
• pp.346-356
• /
• 1992

The extensive Snake River floodplain in Northwest United States has experienced major changes in water channels and vegetation types due to floodings. To detect the change of wetland cover-types for the period of 1985 and 1988, post-classification comparison and image differencing change detection techniques were evaluated using Landsat-5 TM digital data. Differenced infrared-band images indicated better accuracy indices than any visible-band images. A thresholding technique was applied to identify the change and no change categories from the transformed images produced by image differencing. The problems in using different accuracy indices, including the Kappa coefficient of agreement, overall accuracy, producer's accuracy, user's accuracy, and average accuracy(based on both the producer's and user's accuracy approaches) in determining an optimal threshold level, were examined.