• Journal of KIISE
• /
• v.43 no.8
• /
• pp.878-882
• /
• 2016

Automatic generation of captions for an image is a very difficult task, due to the necessity of computer vision and natural language processing technologies. However, this task has many important applications, such as early childhood education, image retrieval, and navigation for blind. In this paper, we describe a Recurrent Neural Network (RNN) model for generating image captions, which takes image features extracted from a Convolutional Neural Network (CNN). We demonstrate that our models produce state of the art results in image caption generation experiments on the Flickr 8K, Flickr 30K, and MS COCO datasets.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_4
• /
• pp.1911-1923
• /
• 2022

The availability of high-resolution satellite images provides precise information without direct observation of the research target. Korea Multi-Purpose Satellite (KOMPSAT), also known as the Arirang satellite, has been developed and utilized for earth observation. The machine learning model was continuously proven as a good classifier in classifying remotely sensed images. This study aimed to compare the performance of the support vector machine (SVM) model in classifying the land cover of the Delaware River port area on high and medium-resolution images. Three optical images, which are KOMPSAT-2, KOMPSAT-3A, and Sentinel-2B, were classified into six land cover classes, including water, road, vegetation, building, vacant, and shadow. The KOMPSAT images are provided by Korea Aerospace Research Institute (KARI), and the Sentinel-2B image was provided by the European Space Agency (ESA). The training samples were manually digitized for each land cover class and considered the reference image. The predicted images were compared to the actual data to obtain the accuracy assessment using a confusion matrix analysis. In addition, the time-consuming training and classifying were recorded to evaluate the model performance. The results showed that the KOMPSAT-3A image has the highest overall accuracy and followed by KOMPSAT-2 and Sentinel-2B results. On the contrary, the model took a long time to classify the higher-resolution image compared to the lower resolution. For that reason, we can conclude that the SVM model performed better in the higher resolution image with the consequence of the longer time-consuming training and classifying data. Thus, this finding might provide consideration for related researchers when selecting satellite imagery for effective and accurate image classification.

• Journal of Integrative Natural Science
• /
• v.12 no.4
• /
• pp.122-126
• /
• 2019

In this paper, we propose a simple and efficient retrieval technique using the feature value of the corner region, which is one of the shape information attributes of images. The proposed algorithm extracts the edges and corner points of the image and rearranges the feature values of the corner regions doubly, and then measures the similarity with the image in the database using the correlation of these feature values as the feature vector. The proposed algorithm is confirmed to be more robust to rotation and size change than the conventional image retrieval method using the corner point.

• Journal of Fisheries and Marine Sciences Education
• /
• v.28 no.2
• /
• pp.376-383
• /
• 2016

In this paper, the blue-green algae monitoring program of applying IoT(Internet of things) technologies is designed and implemented that can check out the status of the river's water quality in real time. The proposed system is to extract the image data from the camera of raspberry pi by an wireless network, and it is analyzed through the HSV color model. We measure the temperature using a DS18B20 1-wire temperature sensor. The extracted information of image data and temperature is then analyzed in C and Python programs for use with Raspberry Pi. The XML data in PHP program is made from the analyzed information and provides Web services. It also allows to refer the XML data using mobile devices.

• Korean Journal of Remote Sensing
• /
• v.23 no.5
• /
• pp.355-363
• /
• 2007

The 3D DWT(The Three Dimensional Discrete Wavelet Transform) scheme is potentially regarded as useful one on analyzing both spatial and spectral information. Nevertheless, few researchers have attempted to process or classified remotely sensed images using the 3D DWT. This study aims to apply the 3D DWT to the land cover classification of optical and SAR(Synthetic Aperture Radar) images. Then, their results are evaluated quantitatively and compared with the results of traditional classification technique. As the experimental results, the 3D DWT shows superior classification results to conventional techniques, especially dealing with the high-resolution imagery and SAR imagery. It is thought that the 3D DWT scheme can be extended to multi-temporal or multi-sensor image classification.

• Journal of Information Science Theory and Practice
• /
• v.7 no.2
• /
• pp.32-39
• /
• 2019

A convolutional neural network (CNN) has been widely used in facial expression recognition (FER) because it can automatically learn discriminative appearance features from an expression image. To make full use of its discriminating capability, this paper suggests a simple but effective method for CNN based FER. Specifically, instead of an original expression image that contains facial appearance only, the expression image with facial geometry visualization is used as input to CNN. In this way, geometric and appearance features could be simultaneously learned, making CNN more discriminative for FER. A simple CNN extension is also presented in this paper, aiming to utilize geometric expression change derived from an expression image sequence. Experimental results on two public datasets (CK+ and MMI) show that CNN using facial geometry visualization clearly outperforms the conventional CNN using facial appearance only.

• Journal of Information Processing Systems
• /
• v.17 no.3
• /
• pp.453-461
• /
• 2021

Many deep learning approaches have been studied for image classification in computer vision. However, there are not enough data to generate accurate models in medical fields, and many datasets are not annotated. This study presents a new method that can use both unlabeled and labeled data. The proposed method is applied to classify cervix images into normal versus cancerous, and we demonstrate the results. First, we use a patch self-supervised learning for training the global context of the image using an unlabeled image dataset. Second, we generate a classifier model by using the transferred knowledge from self-supervised learning. We also apply attention learning to capture the local features of the image. The combined method provides better performance than state-of-the-art approaches in accuracy and sensitivity.

• Journal of The Korean Association of Information Education
• /
• v.24 no.6
• /
• pp.573-584
• /
• 2020

This study is about the development of AI algorithm education program using cognition modeling to positively improve students' image on AI. First, we analyzed the concept of user-based collaborative filtering and developed the education program using the cognition modeling method. We checked the adequacy of program through the expert validity test. Both CVR values for the content development method of cognitive modeling and the developed program showed validity above .80. We applied the developed program to elementary school students in class. The test was conducted using a semantic discrimination to examine changes in students' perception of artificial intelligence before and after. We were able to confirm that the students' AI images were significant positive change in 12 of the 23 words in the adjective pair.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.41 no.5
• /
• pp.796-808
• /
• 2017

This research shows how fashion curation service consumers (both fashion and image consultants) reveal different aspects towards a preferred image type among the correlation of fashion curation service usage motivation and fashion shopping propensity. These preferences also included the purchase intentions for fashion display service products. This study surveyed 300 men and women between the ages of 20-30 who were the main consumers of fashion curation services. 'Convenience, fashion trend, and exploratory' increased purchase intentions for fashion shopping propensity, 'information search for utility, entertainment, and personal expression' increased purchase intentions for fashion curation services and 'brand identity, consumer lifestyle, and product information' increased factors for the fashion curation service preferred image type. Consumer preferences varied according to different fashion curation service image type; however, all consumer group syndicated a difference in fashion curation service actions. For instance, fashion curation service consumers preferred a consumer lifestyle image, convenience, hedonic shopping orientation, and personal expression motivation had a positive influence on product purchase intention. However, the shopping orientation of 'fashion trend, practical information exploration, and entertainment motivation' had an optimistic influence on product purchase intentions for fashion curation service consumers who preferred a brand identity image and a product information image.

• The Korean Journal of Food And Nutrition
• /
• v.25 no.2
• /
• pp.338-347
• /
• 2012

The purpose of this study was to provide basic information regarding desirable body image recognition by examining body image recognition 395 middle school students in the Chungnam area. The average age of the subjects was 13.7 years for boys and 12.6 for girls. Their average height and weight were 165.4 cm and 57.1 kg for boys, and 155.7 cm and 48.8 kg for girls. As for body shape, girls thought that they were overweight more often and wanted to lose weight compared to the boys. Over half of the respondents answered that their weight control efforts were not systematic such ad via professional counseling. Weight control by students was attempted by themselves in order to control their weight by skipping meals. Further, the subjects exhibit dietary behavioral problems such as overeating, skipping meals, unbalanced diet, and eating speed. In conclusion, correct body image recognition is needed and families and schools should make efforts to help students properly control their weight and adopt proper eating habits.