• Journal of Multimedia Information System
• /
• v.9 no.2
• /
• pp.75-86
• /
• 2022

The existing image quality assessment (IQA) datasets have a small number of samples. Some methods based on transfer learning or data augmentation cannot make good use of image quality-related features. A No Reference (NR)-IQA method based on multi-task training and quality awareness is proposed. First, single or multiple distortion types and levels are imposed on the original image, and different strategies are used to augment different types of distortion datasets. With the idea of weak supervision, we use the Full Reference (FR)-IQA methods to obtain the pseudo-score label of the generated image. Then, we combine the classification information of the distortion type, level, and the information of the image quality score. The ResNet50 network is trained in the pre-train stage on the augmented dataset to obtain more quality-aware pre-training weights. Finally, the fine-tuning stage training is performed on the target IQA dataset using the quality-aware weights to predicate the final prediction score. Various experiments designed on the synthetic distortions and authentic distortions datasets (LIVE, CSIQ, TID2013, LIVEC, KonIQ-10K) prove that the proposed method can utilize the image quality-related features better than the method using only single-task training. The extracted quality-aware features improve the accuracy of the model.

• The KIPS Transactions:PartA
• /
• v.11A no.3
• /
• pp.163-172
• /
• 2004

Multimedia applications tend to access their data as a streaming pattern with regular intervals. This characteristic can be utilized in prefetching the multimedia data into cache memory so as to reduce their execution speeds. The reference-prediction prefetch algorithm predicts the memory address that seems to be used in the next time based on the previous history of memory references stored in the prediction reference table. This paper proposes a strategy to manipulate the reference prediction table which contains all of the data reference instructions to scalar and streaming data. We have recognized that the scalar reference instructions do not contribute to the data prefetching algorithm. Therefore, when replacing an element in the reference prediction table, the proposed algorithm preferentially selects the scalar reference instruction before the stream reference instruction. It makes the stream reference instruction to stay for a long time compared to the FIFO replacement policy, and eventually improves the performance of data prefetching.

• Asian-Australasian Journal of Animal Sciences
• /
• v.32 no.2
• /
• pp.159-169
• /
• 2019

Objective: Models for genomic selection assume that the reference population is an unselected population. However, in practice, genotyped individuals, such as progeny-tested bulls, are highly selected, and the reference population is created after preselection. In dairy cattle, the intensity of selection is higher in males than in females, suggesting that cows can be added to the reference population with less bias and loss of accuracy. The objective is to develop formulas applied to any genomic prediction studies or practice with preselected animals as reference population. Methods: We developed formulas for calculating the reliability and bias of genomically enhanced breeding values (GEBV) in the reference population where individuals are preselected on estimated breeding values. Based on the formulas presented, deterministic simulation was conducted by varying heritability, preselection percentage, and the reference population size. Results: The number of bulls equal to a cow regarding the reliability of GEBV was expressed through a simple formula for the reference population consisting of preselected animals. The bull population was vastly superior to the cow population regarding the reliability of GEBV for low-heritability traits. However, the superiority of reliability from the bull reference population over the cow population decreased as heritability increased. Bias was greater for bulls than cows. Bias and reduction in reliability of GEBV due to preselection was alleviated by expanding reference population. Conclusion: Cows are easier in expanding reference population size compared with bulls and alleviate bias and reduction in reliability of GEBV of bulls which are highly preselected than cows by expanding the cow reference population.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.27 no.2
• /
• pp.239-246
• /
• 2003

This paper proposes engineering estimation equations of elastic-plastic J and COD fur axial through-wall cracked pipes under internal pressure. Based on detailed 3-D FE results based on deformation plasticity, the plastic influence functions for fully plastic J and COD solutions are tabulated as a function of the mean radius-to-thickness ratio, the normalized crack length. and the strain hardening. Based on these results, the GE/EPRI-type J and COD estimation equations are proposed and validated against the 3-D FE results based on deformation plasticity. For more general application to general stress-strain laws or to complex loading, the developed GE/EPRI-type solutions are re-formulated based on the reference stress concept. Such a reformulation provides simpler equations for J and COD, which are then further extended to combined internal pressure and bending. The proposed reference stress based J and COD estimation equations are compared with elastic-plastic 3-D FE results using actual stress-strain data for Type 316 stainless steels. The FE results for both internal pressure cases and combined internal pressure and bending cases compare very well with the proposed J and COD estimations.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.26 no.6
• /
• pp.617-624
• /
• 2008

In general, RTK-GPS(Real Time Kinematic GPS) based on the single reference station is able to determinate the high accurate position of rover on the spot using error correction information of transmitted carrier phase from the base station via wireless modem. However, single reference station method has some weak points to decrease positioning accuracy because it must be obtained carrier phase from the each satellite continuously, allowed to transmit without obstacle and limited to short base line distance between base and rover station. This paper aims to attempt network based GPS carrier phase differential positioning using three multi reference stations to overcome the method of single reference station and RTK network is realized by real time monitoring program with Visual C++. The optimum error correction value of three multi reference stations by RTK networking is selected automatically to correct the position of rover station. In this paper, this algorithm is applied to determine sea water level using GPS buoy, and the accuracy results of water level change were analyzed and compared with each other using single and multi reference stations.

• Journal of Korean Society of Archives and Records Management
• /
• v.22 no.3
• /
• pp.103-124
• /
• 2022

This study aims to suggest the use-based policies for archival reference service by investigating the actual archival uses and services of the National Archives of Korea (NAK). Considering that there is little information on the overall archival reference service by NAK, an exploratory research method is adopted to understand the current status and issues of the service. In this study, the use of archival records in NAK is investigated from statistical data, and service cases by record type are collected through interviews with archival service professionals of NAK. Then, the success and risk factors of the archival reference service are analyzed in three categories of human resources (reference archivists' competency), tools (finding aids), and records (appropriateness and sufficiency). Based on these analyses, archival reference service policy directions are proposed to strengthen the two different archival roles as "arsenal of rights protection" and "academic information provider."

• The korean journal of orthodontics
• /
• v.49 no.2
• /
• pp.97-105
• /
• 2019

Objective: This study was performed to evaluate the accuracy and reliability of a newly designed method to achieve mandibular dental model superimposition, using voxel-based cone-beam computed tomography (CBCT) registration. Methods: Fourteen dry cadaveric mandibles and six teeth extracted from patients with severe periodontitis were used to establish 14 orthodontic tooth-movement models. The protocol consisted of two steps: in the first step, voxel-based CBCT mandible superimposition was performed; the reference comprised the external portion of the symphysis, extending to the first molar. The laser-scanned dental model image was then integrated with the CBCT image to achieve mandibular dental model superimposition. The entire process required approximately 10 minutes. Six landmarks were assigned to the teeth to measure tooth displacement, using tooth displacement on the superimposed laser-scanned mandibles as the reference standard. Accuracy was evaluated by comparing differences in tooth displacement based on the method and the reference standard. Two observers performed superimposition to evaluate reliability. Results: For three-dimensional tooth displacements, the differences between the method and the reference standard were not significant in the molar, premolar, or incisor groups (p > 0.05). The intraclass correlation coefficients for the inter- and intra-observer reliabilities of all measurements were > 0.92. Conclusions: Our method of mandibular dental model superimposition based on voxel registration is accurate, reliable, and can be performed within a reasonable period of time in vitro, demonstrating a potential for use in orthodontic patients.

• The Journal of Advanced Prosthodontics
• /
• v.12 no.6
• /
• pp.361-368
• /
• 2020

PURPOSE. The present study aimed to evaluate the accuracy of a desktop scanner and intraoral scanners based on the volumetric dimensions of a complete arch. MATERIALS AND METHODS. Seven reference models were fabricated based on the volumetric dimensions of complete arch (70%, 80%, 90%, 100%, 110%, 120%, and 130%). The reference models were digitized using an industrial scanner (Solutionix C500; MEDIT) for the fabrication of a computer-aided design (CAD) reference model (CRM). The reference models were digitized using three intraoral scanners (CS3600, Trios3, and i500) and one desktop scanner (E1) to fabricate a CAD test model (CTM). CRM and CTM were then superimposed using inspection software, and 3D analysis was conducted. For statistical analysis, one-way analysis of variance was used to verify the difference in accuracy based on the volumetric dimensions of the complete arch and the accuracy based on the scanners, and the differences among the groups were analyzed using the Tukey HSD test as a post-hoc test (α=.05). RESULTS. The three different scanners showed a significant difference in accuracy based on the volumetric dimensions of the complete arch (P<.05), but the desktop scanner did not show a significant difference in accuracy based on the volumetric dimensions of the complete arch (P=.808). CONCLUSION. The accuracy of the intraoral scanners was dependent on the volumetric dimensions of the complete arch, but the volumetric dimensions of the complete arch had no effect on the accuracy of the desktop scanner. Additionally, depending on the type of intraoral scanners, the accuracy differed according to the volumetric dimensions of the complete arch.

• Journal of Broadcast Engineering
• /
• v.19 no.2
• /
• pp.250-256
• /
• 2014

Non-reference P frame coding is used to enhance coding efficiency in low-delay encoding configuration of Internet Video Coding (IVC), which is being standardized as a royalty-free video codec in MPEG. The existing method of non-reference P frame coding which was adopted in the reference Test Model of IVC (ITM) 4.0 adaptively applies a non-reference P frame with a fixed coding structure based on the magnitude of motion vectors (MVs), however, which unexpectedly degrades the coding efficiency for some sequences. In this paper, the existing non-reference P frame coding is improved by changing non-reference P frame coding structure and applying a new adaptive method using the ratio of the amount of generated bits of non-reference frames to that of reference frames as well as MVs. Experimental results show that the proposed non-reference P frame coding gives 6.6% BD-rate bit saving in average over ITM 7.0.

• Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference
• /
• 2004.04a
• /
• pp.91-96
• /
• 2004

More recently, multiple reference based RTK GPS techniques(VRS : Virtual Reference System) are becoming increasingly important for many precise GPS applications in many countries to overcome the constrained distance limitations of standard RTK systems. The precision of the position solutions of the rover receiver is closely connected with that of the corresponding virtual reference points(VRPs). The objective of this paper is to investigate the accuracy and performance of the VRPs on the test network for static positioning in post processing mode. For this, some VRPs at the inside and outside of test network have made in post mode, and the overall analysis results were presented by comparing the solution for the VRPs from the existed GPS reference station with the true values of the coordinates used to produce the observation data. The results show the reasonable accuracies of VRPs in the network area by using the VRS concept in post mode.