• Journal of Broadcast Engineering
• /
• v.27 no.3
• /
• pp.273-282
• /
• 2022

MPEG-VCM(Video Coding for Machine) aims to standardize video codec for machines. VCM provides data sets and anchors, which provide reference data for comparison, for several machine vision tasks including object detection, object segmentation, and object tracking. The evaluation template can be used to compare compression and machine vision task performance between anchor data and various proposed video codecs. However, performance comparison is carried out separately for each machine vision task, and information related to performance evaluation of multiple machine vision tasks on a single bitstream is not provided currently. In this paper, we propose a performance evaluation method of a video codec for AI-based multi-tasks. Based on bits per pixel (BPP), which is the measure of a single bitstream size, and mean average precision(mAP), which is the accuracy measure of each task, we define three criteria for multi-task performance evaluation such as arithmetic average, weighted average, and harmonic average, and to calculate the multi-tasks performance results based on the mAP values. In addition, as the dynamic range of mAP may very different from task to task, performance results for multi-tasks are calculated and evaluated based on the normalized mAP in order to prevent a problem that would be happened because of the dynamic range.

• 제어로봇시스템학회:학술대회논문집
• /
• 1990.10a
• /
• pp.423-428
• /
• 1990

A mixed IP formulation is presented which concurrently routes and sequences the tasks on the MRS, reflecting the flexible characteristics. As a preliminary work for the performance evaluation of the MRS, a two robot system working on independent tasks is studied. Models for three types of the system capable of simple error recovery function are established using queueing model, and performances are evaluated and analysed.

• Communications for Statistical Applications and Methods
• /
• v.19 no.4
• /
• pp.571-577
• /
• 2012

Modules that consist of software are respectively coded in the early development phase and the modules are unified as a software. After unification, the software is repeatedly tested with a given taskset (the set of module tasks that are tested simultaneously) until a required performance level is satisfied. In this paper, we expand the one-module software debugging model of Jang and Lee (2011) to a multi-module debugging model and derive the taskset completion probability and the mean of the completed tasksets under the assumption that the processing times of module tasks given in a taskset are mutually dependent.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.15 no.9
• /
• pp.5652-5659
• /
• 2014

The purpose of this study was to introduce a software task processing evaluation model that considers the following situations: i) a software system is integratedly composed of several number of modules, ii) each modules has its corresponding module task, iii) all module tasks are tested simultaneously, and iv) the processing times of the module tasks are mutually dependent. The software task completion probability with the module dependency was derived using the joint distribution function of Farlie [11]. The results showed that the task completion probability of software increases with increasing module dependency parameter.

• Proceedings of the KSR Conference
• /
• 2003.05a
• /
• pp.519-524
• /
• 2003

Systems engineering test and evaluation technology development program for (Korea) Express Tilting Train(TTX) system in progress is a national large system development program that is not only a large-size and complex but also multi-disciplinary in nature. For the TTX test and evaluation technology development, verification requirements that could verify system function and performance, should be derived from system specification. Hereafter, this could be listed as items of test. System engineering process establish traceability between items of test and system requirements. These tasks could be accomplished by the schema. Using the Core, a system engineering tool, TTX program could be database and similar other system could be developed effectively and efficiency

• Smart Media Journal
• /
• v.12 no.6
• /
• pp.35-40
• /
• 2023

In this study paper, we examine pre-trained language models used in Multi-Goal Conversational Recommender Systems (MG-CRS), comparing and analyzing their performances of various pre-trained language models. Specifically, we investigates the impact of the sizes of language models on the performance of MG-CRS. The study targets three types of language models - of BERT, GPT2, and BART, and measures and compares their accuracy in two tasks of 'type prediction' and 'topic prediction' on the MG-CRS dataset, DuRecDial 2.0. Experimental results show that all models demonstrated excellent performance in the type prediction task, but there were notable provide significant performance differences in performance depending on among the models or based on their sizes in the topic prediction task. Based on these findings, the study provides directions for improving the performance of MG-CRS.

• The Korean Journal of Applied Statistics
• /
• v.21 no.6
• /
• pp.997-1006
• /
• 2008

This paper proposes a new model by combining an infinite-server queueing model for multi-task processing software system with a perfect debugging model based on Markov process with two types of faults suggested by Lee et al. (2001). We apply this model for module and integration testing in the testing process. Also, we compute several measure, such as the expected number of tasks whose processes can be completed and the task completion probability are investigated under the proposed model.

• Journal of Institute of Control, Robotics and Systems
• /
• v.4 no.1
• /
• pp.113-122
• /
• 1998

This paper presents an implementation method of Profibus interface software using FMS(Fieldbus Message Specification). The Profibus interface software is implemented on PC which is widely used as an industrial computer as well as a commercial embedded controller called IUC(Intelligent Universal Controller). In order to enable the Profibus interface software to handle many application tasks and communication services, two kinds of real-time/multi-tasking operating system, OS-9 and CTask, are utilized. We also develop an experimental model of Profibus-based automation system, and evaluate the performance of the Profibus network. Through experiments, the user layer level message latency is measured with respect to the change of message length, message generation interval and TRT(Target Rotation Time). The results of experiment are compared with those of a simulation model which comprises only the physical and data link layers of Profibus. The results of this study shows that the message latency in the user layer level occupies fairly a large part of the total message latency.

• The Journal of Korea Robotics Society
• /
• v.19 no.1
• /
• pp.130-138
• /
• 2024

High-precision 3D Object Detection is a crucial component within autonomous driving systems, with far-reaching implications for subsequent tasks like multi-object tracking and path planning. In this paper, we propose a novel approach designed to enhance the performance of 3D Object Detection, especially in heading angle estimation by employing a moving object segmentation technique. Our method starts with extracting point-wise moving labels via a process of moving object segmentation. Subsequently, these labels are integrated into the LiDAR Pointcloud data and integrated data is used as inputs for 3D Object Detection. We conducted an extensive evaluation of our approach using the KITTI-road dataset and achieved notably superior performance, particularly in terms of AOS, a pivotal metric for assessing the precision of 3D Object Detection. Our findings not only underscore the positive impact of our proposed method on the advancement of detection performance in lidar-based 3D Object Detection methods, but also suggest substantial potential in augmenting the overall perception task capabilities of autonomous driving systems.

• Journal of the Korea Safety Management & Science
• /
• v.23 no.3
• /
• pp.97-102
• /
• 2021

This study developed a scenario to understand the reaction rate and operational time according to RTI value of rate of rise detector in each type in case of fire mattress. In the results of analyzing the reaction rate and operational time of detector in each scenario, in case when installing a single detector, the elevated temperature per minute was raised to 8℃/min ~ 9℃/min. In case when installing two detectors, it was raised to 9℃/min ~ 10℃/min. In case when installing three detectors, it was raised to 10℃/min. The horizontal distance between detector and mattress was 1.8m~2.5m. Whenever the number of detectors was increased, the horizontal distance was decreased. The operational time of detector was within maximum 540 seconds and minimum 420 seconds. As the research tasks in the future, there should be the researches on the effects of reaction rate of detector on the evacuation in case of fire through the result value of RSET by setting up the latency until the detector operates, and the researches on the safety by understanding if the operational time of detector is suitable for the evaluation standard of performance-centered design.