• Journal of KIISE
• /
• v.43 no.1
• /
• pp.47-53
• /
• 2016

While the Internet develops rapidly, a huge amount of image data collected from smart phones, digital cameras and black boxes are being shared through social media sites. Generally, social images are handled by tagging them with information. Due to the ease of sharing multimedia and the explosive increase in the amount of tag information, it may be considered too much hassle by some users to put the tags on images. Image retrieval is likely to be less accurate when tags are absent or mislabeled. In this paper, we suggest a method of extracting tags from social images by using image content. In this method, CNN(Convolutional Neural Network) is trained using ImageNet images with labels in the training set, and it extracts labels from instagram images. We use the extracted labels for automatic image tagging. The experimental results show that the accuracy is higher than that of instagram retrievals.

• Journal of KIISE
• /
• v.42 no.5
• /
• pp.637-641
• /
• 2015

Due to the recent increasement of user created contents like SNS, blog posts, and so on, broadcast contents are actively re-construction by its users. Especially, on some genres like drama, movie, various information from cars and film sites to clothes and watches in a content is spreaded out to other users through blog postings. Since such information can be an additional information for the content, they can be used for providing high-quality broadcast services. For this purpose, in this paper, we propose timeline tag cloud generation method for broadcasting contents. In the proposed method, blog postings on the target contents are first gathered and then, images and words around images are extracted from a blog post as a tag set. An extracted tag set is tagged on a specific timeline of the target content. In experiments, to prove the efficiency of the proposed method, we evaluated the performances of the proposed image matching and tag cloud generation methods.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.6 no.3
• /
• pp.81-88
• /
• 1996

Cluster analysis is based on partitioning a collection of data points into a number of clusters, where the data points in side a cluster have a certain degree of similarity and it is a fundamental process of data analysis. So, it has been playing an important role in solving many problems in pattern recognition and image processing. For these many clustering algorithms depending on distance criteria have been developed and fuzzy set theory has been introduced to reflect the description of real data, where boundaries might be fuzzy. If fuzzy cluster analysis is tomake a significant contribution to engineering applications, much more attention must be paid to fundamental questions of cluster validity problem which is how well it has identified the structure that is present in the data. Several validity functionals such as partition coefficient, claasification entropy and proportion exponent, have been used for measuring validity mathematically. But the issue of cluster validity involves complex aspects, it is difficult to measure it with one measuring function as the conventional study. In this paper, we propose four performance indices and the way to measure the quality of clustering formed by given learning strategy.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.12 no.5
• /
• pp.937-941
• /
• 2008

People who have hearing difficulties use sign language as the most important communication method, and they can broaden personal relations and manage their everyday lives without inconvenience through sign language. But they suffer from absence of interpolation between normal people and people who have hearing difficulties in increasing video chatting or video communication services by recent growth of internet communication. In this paper, we proposed a sign language recognition method in order to solve such a problem. In the proposed method, regions of two hands are extracted by tracking of two hands using RGB, YUV and HSI color information from a sign language image acquired from a video camera and by removing noise in the segmented images. The extracted regions of two hands are teamed and recognized by ART2 algorithm that is robust for noise and damage. In the experiment by the proposed method and images of finger number from 1 to 10, we verified the proposed method recognize the numbers efficiently.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.5
• /
• pp.1004-1012
• /
• 2007

In this paper, we proposed a new car license plate recognition method using an unsupervised ART2 algorithm with HSI color model. The proposed method consists of two main modules; extracting plate area from a vehicle image and recognizing the characters in the plate after that. To extract plate area, hue(H) component of HSI color model is used, and the sub-area containing characters is acquired using modified fuzzy binarization method. Each character is further divided by a 4-directional edge tracking algorithm. To recognize the separated characters, noise-robust ART2 algorithm is employed. When the proposed algorithm is applied to recognize license plate characters, the extraction rate is better than that of existing RGB model and the overall recognition rate is about 97.4%.

• Journal of Korea Society of Industrial Information Systems
• /
• v.26 no.5
• /
• pp.1-9
• /
• 2021

Children's drawings are widely used in the diagnosis of children's psychology as a means of expressing inner feelings. This paper proposes a children's drawings-based object detection algorithm applicable to children's psychology analysis. First, the sketch area from the picture was extracted and the data labeling process was also performed. Then, we trained and evaluated a Faster R-CNN based object detection model using the labeled datasets. Based on the detection results, information about the drawing's area, position, or color histogram is calculated to analyze primitive information about the drawings quickly and easily. The results of this paper show that Artificial Intelligence-based object detection algorithms were helpful in terms of psychological analysis using children's drawings.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.21 no.5
• /
• pp.197-202
• /
• 2021

Biometric technology using finger veins is receiving a lot of attention due to its high security, convenience and accuracy. And the recent development of deep learning technology has improved the processing speed and accuracy for authentication. However, the training data is a subset of real data not in a certain order or method and the results are not constant. so the amount of data and the complexity of the artificial neural network must be considered. In this paper, the deep learning model of Inception-Resnet-v2 was used to improve the high accuracy of the finger vein recognizer and the performance of the authentication system, We compared and analyzed the performance of the deep learning model of DenseNet-201. The simulations used data from MMCBNU_6000 of Jeonbuk National University and finger vein images taken directly. There is no preprocessing for the image in the finger vein authentication system, and the results are checked through EER.

• Journal of Broadcast Engineering
• /
• v.26 no.5
• /
• pp.608-620
• /
• 2021

Recent deep learning-based face super-resolution (FSR) works showed significant performances by utilizing facial prior knowledge such as facial landmark and dictionary that reflects structural or semantic characteristics of the human face. However, most of these methods require additional processing time and memory. To solve this issue, this paper propose an efficient FSR models using knowledge distillation techniques. The intermediate features of teacher network which contains dictionary information based on major face regions are transferred to the student through adversarial multi-scale features distillation. Experimental results show that the proposed model is superior to other SR methods, and its effectiveness compare to teacher model.

• Journal of the Korea Convergence Society
• /
• v.12 no.9
• /
• pp.57-63
• /
• 2021

This study dissected and three-dimensionally (3D) scanned the facial muscles of Korean adult cadavers, created a three-dimensional model with realistic facial muscle shapes, and reproduced facial expressions to provide educational materials to allow the 3D observation of the complex movements of cadaver facial muscles. Using the cadavers' anatomical photo data, 3D modeling of facial muscles was performed. We produced models describing four different expressions, namely sad, happy, surprised, and angry. We confirmed the complex action of the 3D cadaver facial muscles when making various facial expressions. Although the results of this study cannot confirm the individual functions of facial muscles quantitatively, we were able to observe the realistic shape of the cadavers' facial muscles, and produce models that would show different expressions depending on the actions performed. The data from this study may be used as educational materials when studying the anatomy of facial muscles.

• The Journal of Korean Institute of Next Generation Computing
• /
• v.15 no.6
• /
• pp.7-14
• /
• 2019

The automatic monitoring and detection of crowd behavior in the surveillance videos has obtained significant attention in the field of computer vision due to its vast applications such as security, safety and protection of assets etc. Also, the field of crowd analysis is growing upwards in the research community. For this purpose, it is very necessary to detect and analyze the crowd behavior. In this paper, we proposed a deep learning-based method which detects abnormal activities in surveillance cameras installed in a smart city. A fine-tuned VGG-16 model is trained on publicly available benchmark crowd dataset and is tested on real-time streaming. The CCTV camera captures the video stream, when abnormal activity is detected, an alert is generated and is sent to the nearest police station to take immediate action before further loss. We experimentally have proven that the proposed method outperforms over the existing state-of-the-art techniques.