• Journal of Korea Multimedia Society
• /
• v.7 no.5
• /
• pp.617-632
• /
• 2004

Video Data includes moving objects which change spatial positions as time goes by. In this paper, we propose a new modeling method for a moving object contained in the video data. In order to effectively retrieve moving objects, the proposed modeling method represents the spatial position and the size of a moving object. It also represents the visual features and the trajectory by considering direction, distance and speed or moving objects as time goes by. Therefore, It allows various types of retrieval such as visual feature based similarity retrieval, distance based similarity retrieval and trajectory based similarity retrieval and their mixed type of weighted retrieval.

• Journal of Korean Society for Quality Management
• /
• v.20 no.2
• /
• pp.109-117
• /
• 1992

An inspection task largely can be seen as a job divided up into a series of visual search and classification subtasks. In these subtasks, an Inspector must performs to compare the standard references proposed in visual environments and recalled in his memory with the visual stimuli to be inspected. It means that the judgement of similarity should be demanded on inspection tasks. Therefore, the inspector's ability for the judgement of similarity and the difference similarity between inspection materials are important factors to effect on performances in inspection tasks. In this paper, to analysis the effect of these factors on inspection time, an inspection task is designed and suggested by means of computer simulator. Especially, the skin conductance responses(SCR) of subjects are measured to evaluate the complexity of tasks due to the difference of similarity between materials. In the results of experiment, the more similar or different the difference of similarity between materials is, the shorter the inspection time is because of the reduction of task complexity. And, When the inspector's cognition for similarity between materials is consistanct, the inpsection time is improved. Concludingly, the consistency of reponses for similarity judgement becomes a measurement to present the performance levels. And the information of inspection time that due to the difference of similarity between materials must be considered in planning and scheduling inspection tasks.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.7
• /
• pp.57-64
• /
• 2020

In this paper, we propose an scalable re-detection for correlation filter in visual tracking. In real world, there are lots of target disappearances and reappearances during tracking, thus failure detection and re-detection methods are needed. One of the important point for re-detection is that a searching area must be large enough to find the missing target. For robust visual tracking, we adopt kernelized correlation filter as a baseline. Correlation filters have been extensively studied for visual object tracking in recent years. However conventional correlation filters detect the target in the same size area with the trained filter which is only 2 to 3 times larger than the target. When the target is disappeared for a long time, we need to search a wide area to re-detect the target. Proposed algorithm can search the target in a scalable area, hence the searching area is expanded by 2% in every frame from the target loss. Four datasets are used for experiments and both qualitative and quantitative results are shown in this paper. Our algorithm succeed the target re-detection in challenging datasets while conventional correlation filter fails.

• Journal of Broadcast Engineering
• /
• v.18 no.5
• /
• pp.691-701
• /
• 2013

In this paper, an early termination algorithm for merge process is proposed to reduce the computational complexity in High Efficiency Video Coding (HEVC) encoder. In the HEVC, the same candidate modes from merge candidate list (MCL) are shared to predict a merge or merge SKIP mode. This search process is performed by the number of the obtained candidates for the both of the merge and SKIP modes. This may cause some redundant search operations. To reduce this redundant search operation, we employ the neighboring blocks which have been encoded in prior, to check on the contextual information. In this study, the spatial, temporal and depth neighboring blocks have been considered to compute a correlation information. With this correlation information, an early termination algorithm for merge process is suggested. When all modes of neighboring blocks are SKIP modes, then the merge process performs only SKIP mode. Otherwise, usual merge process of HEVC is performed Through experimental results, the proposed method achieves a time-saving factor of about 21.25% on average with small loss of BD-rate, when comparing to the original HM 10.0 encoder.

• Journal of the Korean Society for Library and Information Science
• /
• v.50 no.1
• /
• pp.417-435
• /
• 2016

Relevance judgment is important in terms of improving the effectiveness of information retrieval systems, and it has been dominant for users to search and use images utilizing internet and digital technologies. However, in the field of image retrieval, there have been only a few studies in terms of identifying relevance criteria. The purpose of this study aims to identify and characterize the non-relevance criteria from the failed image searches. In order to achieve the purpose of this study, a total of 135 participants were recruited and a total of 1,452 criteria items were collected for this study. Analyses and identification on the data set found thirteen criteria such as 'topicality', 'visual content', 'accuracy', 'visual feature', 'completeness', 'appeal to user', 'focal point', 'bibliographic information', 'impression', 'posture', 'face feature', 'novelty', and 'time frame'. Among these criteria, 'visual content' and 'focal point' were introduced in this current study, while 'action' criterion identified in previous studies was not shown in this current study. When image needs and image uses are analyzed with these criteria, there are distinctive differences depending on different image needs and uses.

• Korean Institute of Interior Design Journal
• /
• v.25 no.5
• /
• pp.71-81
• /
• 2016

This research has been carried out with the objects of sporting goods shops to find out what structure of those shops raises more interest from customers. The tracking eyes on the objects which are the same but seen to have different structures has revealed the followings. Customers' visual appreciation of Reverse left/right Images (11.1) was found to be higher than that of Original Images (10.6). Furthermore, the reverse left/right image of the space also was found to attract more interest from customers, which led them to have longer observation. The below is about the interpretation of the spatial exploration by observation time and the appreciation of its visual content in line with the experiment objects of selling spaces. The longer the space was observed, (1)the higher the expansive searching of space was, (2)the more spots were observed as if they did not know what to see after they first observed at early hours, (3)later (in the time range of 64~73 seconds) they came to look at the spots in which they got interested, (4)and then again they suddenly got lost what to see. When the change of observation characteristics by time range is reviewed, it can be seen that the searching of original images is changed from Divergent Feature to Convergent Feature when the observation time increases from the early stage of observation to the later. On the contrary, the reverse left/right images were found to have the opposite searching features, that is, from convergent exploration to divergent exploration. These findings show that the reverse left/right images of the sporting goods shops, which were the experiment objects, have more factors attracting customers' attention and interest and that it is the very shop-structure which makes customers have better visual appreciation of those shops.

• Korean Journal of Cognitive Science
• /
• v.32 no.3
• /
• pp.117-139
• /
• 2021

The way of change detection in which presence of a different item is determined between memory and test arrays with a brief in-between time interval resembles how visual search is done considering that the different item is searched upon the onset of a test array being compared against the items in memory. According to the resemblance, the present study examined whether varying the probability of change occurrence in a visual short-term memory-based change detection task can influence the aspect of response-decision making (i.e., change prevalence effect). The simple-feature change detection task in the study consisted of a set of four colored boxes followed by another set of four colored boxes between which the participants determined presence or absence of a color change from one box to the other. The change prevalence was varied to 20, 50, or 80% in terms of change occurrences in total trials, and their change detection errors, detection sensitivity, and their subsequent RTs were analyzed. The analyses revealed that as the change prevalence increased, false alarms became more frequent while misses became less frequent, along with delayed correct-rejection responses. The observed change prevalence effect looks very similar to the target prevalence effect varying according to probability of target occurrence in visual search tasks, indicating that the background principles deriving these two effects may resemble each other.

• Journal of the Korea Computer Graphics Society
• /
• v.23 no.4
• /
• pp.21-29
• /
• 2017

Previous visual analytics researches has focused on reducing the uncertainty of predicted results using a variety of interactive visual data exploration techniques. The main purpose of the interactive search technique is to reduce the quality difference of the predicted results according to the level of the decision maker by understanding the relationship between the variables and choosing the appropriate model to predict the unknown variables. However, it is difficult to create a predictive model which forecast time series data whose overall trends is unknown such as youth physical growth data. In this paper, we pro pose a novel predictive analysis technique to forecast the physical growth value in small pieces of time series data with un certain trends. This model estimates the distribution of data at a particular point in time. We also propose a visual analytics system that minimizes the possible uncertainties in predictive modeling process.

• Journal of Broadcast Engineering
• /
• v.12 no.4
• /
• pp.303-310
• /
• 2007

Motion Estimation (ME) is a core part of most Video compression systems since it affects directly the output video quality and the encoding time. The most basic method of ME, Full Search (FS) gives the highest visual quality but also has the problem of significant computational load. To solve this problem, many fast algorithm has been proposed. Among them, MVFAST and PMVFAST show impressive results in video quality and the computational load by using the correlation between motion vectors of adjacent blocks. In particular, PMVFAST reduces search points dramatically and also gives very high video quality by using the median predictor. In this paper, we propose a new algorithm that uses the redefined median predictor which reduces the number of search points and yields a high visual quality by reducing the number of thresholds and early termination conditions.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.12
• /
• pp.3976-3990
• /
• 2022

Currently, cross-attention is widely used in Siamese trackers to replace traditional correlation operations for feature fusion between template and search region. The former can establish a similar relationship between the target and the search region better than the latter for robust visual object tracking. But existing trackers using cross-attention only focus on rich semantic information of high-level features, while ignoring the appearance information contained in low-level features, which makes trackers vulnerable to interference from similar objects. In this paper, we propose a Multi-level Cross-attention Siamese network(MCSiam) to aggregate the semantic information and appearance information at the same time. Specifically, a multi-level cross-attention module is designed to fuse the multi-layer features extracted from the backbone, which integrate different levels of the template and search region features, so that the rich appearance information and semantic information can be used to carry out the tracking task simultaneously. In addition, before cross-attention, a target-aware module is introduced to enhance the target feature and alleviate interference, which makes the multi-level cross-attention module more efficient to fuse the information of the target and the search region. We test the MCSiam on four tracking benchmarks and the result show that the proposed tracker achieves comparable performance to the state-of-the-art trackers.