• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.11
• /
• pp.1437-1444
• /
• 2020

With the development of computer vision systems, many advances have been made in the fields of surveillance, biometrics, medical imaging, and autonomous driving. In the field of autonomous driving, in particular, the object detection technique using deep learning are widely used, and the paved road detection is a particularly crucial problem. Unlike the ROI detection algorithm used in general object detection, the structure of paved road in the image is heterogeneous, so the ROI-based object recognition architecture is not available. In this paper, we propose a deep neural network architecture for atypical paved road detection using Semantic segmentation network. In addition, we introduce the multi-scale semantic segmentation network, which is a network architecture specialized to the paved road detection. We demonstrate that the performance is significantly improved by the proposed method.

• Journal of Information Processing Systems
• /
• v.14 no.6
• /
• pp.1318-1330
• /
• 2018

Video captioning refers to the process of extracting features from a video and generating video captions using the extracted features. This paper introduces a deep neural network model and its learning method for effective video captioning. In this study, visual features as well as semantic features, which effectively express the video, are also used. The visual features of the video are extracted using convolutional neural networks, such as C3D and ResNet, while the semantic features are extracted using a semantic feature extraction network proposed in this paper. Further, an attention-based caption generation network is proposed for effective generation of video captions using the extracted features. The performance and effectiveness of the proposed model is verified through various experiments using two large-scale video benchmarks such as the Microsoft Video Description (MSVD) and the Microsoft Research Video-To-Text (MSR-VTT).

• Journal of the Korean Society of Clothing and Textiles
• /
• v.25 no.3
• /
• pp.560-568
• /
• 2001

The aim of the present study was to measure intra-individual consistency in clothing and fabric evaluation and to examine its relation to the ratings. A sample of 93 female and 97 male university students rated clothing of 4 styles of daytime wear and 2 fabrics on 15 pairs of polar adjectives twice in 7-days interval. Correlation coefficients between the two ratings for each subject, intra-individual consistency in the evaluation, ranged from -0.12 to 0.89 and mean coefficient was 0.63 of female and -0.01 to 0.78 and mean coefficient was 0.54 of male. Based on the coefficients, the subjects were classified into three groups: high, medium, and low intra-individual consistency. Analysis of variance of mean ratings by the three groups revealed that significant difference existed in 24% of female and 23% of male in 90 combinations of 6 clothing and 15 semantic differential scales. Female of subjects with high intra-individual consistency were most likely definite to evaluate clothing, whereas the ones with low were least. But male subjects were not definite. Mean correlation coefficients for style evaluation subscales of female was 0.39, but male was 0.44. Among the semantic differential scales, high stability in the two ratings was observed for the synthetic clothing evaluation. Correlation coefficients for each clothing obtained from the mean score of the subjects in each semantics differential scale were around 0.98, including that the mean scores of the subjects in each scale could yield excellent stability in clothing evaluation.

• Journal of The Korean Association of Information Education
• /
• v.21 no.5
• /
• pp.527-535
• /
• 2017

In this study, we analyzed the image of artificial intelligence recognized by elementary students using semantic differential scale. First, we extracted 23 pairs of image adjectives related to perception of artificial intelligence. Adjectives were classified into three types related to recognition, emotion and ability and 827 elementary students were examined. Image factors were classified into four factors: convenience, technological progress, human-friendliness, and concern. As a result, they showed a clear image that artificial intelligence is clever, new, and complex but exciting. In comparison with variables, female students, coding experience and older students thought that artificial intelligence was more human-friendly and technological progressive.

• Journal of the Korean Institute of Rural Architecture
• /
• v.25 no.4
• /
• pp.1-8
• /
• 2023

This study is meaningful in figuring out how much the Korean people's awareness of hanok has increased even though interest in hanok has also increased due to the Korean Wave craze. Therefore, with respect to the exterior of hanok, which is visually recognized first, the level of experts and ordinary people is grasped through a semantic discrimination scale, and the degree of visual recognition is to be investigated centering on the color image of hanok buildings. This is the process of thinking about how the Korean image should be reflected in the design, and we want to suggest the direction that modern hanok should go. The study compared and analyzed the difference in visual color based on the elevation of the hanok using a 7-point and 5-point scale method for the general public and experts, and utilized the IRI adjective vocabulary scale and the color matching image scale to construct new hanoks with insufficient differences in appearance and shape. It can be applied to design and image preservation and construction of existing hanok.

• ETRI Journal
• /
• v.45 no.5
• /
• pp.822-835
• /
• 2023

Recent semantic segmentation frameworks usually combine low-level and high-level context information to achieve improved performance. In addition, postlevel context information is also considered. In this study, we present a Context ReFinement Network (CRFNet) and its training method to improve the semantic predictions of segmentation models of the encoder-decoder structure. Our study is based on postprocessing, which directly considers the relationship between spatially neighboring pixels of a label map, such as Markov and conditional random fields. CRFNet comprises two modules: a refiner and a combiner that, respectively, refine the context information from the output features of the conventional semantic segmentation network model and combine the refined features with the intermediate features from the decoding process of the segmentation model to produce the final output. To train CRFNet to refine the semantic predictions more accurately, we proposed a sequential training scheme. Using various backbone networks (ENet, ERFNet, and HyperSeg), we extensively evaluated our model on three large-scale, real-world datasets to demonstrate the effectiveness of our approach.

• Journal of Information Processing Systems
• /
• v.17 no.5
• /
• pp.892-904
• /
• 2021

An image semantic segmentation model is proposed based on improved ENet network in order to achieve the low accuracy of image semantic segmentation in complex environment. Firstly, this paper performs pruning and convolution optimization operations on the ENet network. That is, the network structure is reasonably adjusted for better results in image segmentation by reducing the convolution operation in the decoder and proposing the bottleneck convolution structure. Squeeze-and-excitation (SE) module is then integrated into the optimized ENet network. Small-scale targets see improvement in segmentation accuracy via automatic learning of the importance of each feature channel. Finally, the experiment was verified on the public dataset. This method outperforms the existing comparison methods in mean pixel accuracy (MPA) and mean intersection over union (MIOU) values. And in a short running time, the accuracy of the segmentation and the efficiency of the operation are guaranteed.

• Proceedings of the Korea Society of Design Studies Conference
• /
• 1998.05a
• /
• pp.11-12
• /
• 1998

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.12
• /
• pp.1777-1783
• /
• 2021

The use of drones is gradually increasing for the efficient maintenance of large-scale renewable energy power generation complexes. For a long time, photovoltaic panels have been photographed with drones to manage panel loss and contamination. Various approaches using artificial intelligence are being tried for efficient maintenance of large-scale photovoltaic complexes. Recently, semantic segmentation-based application techniques have been developed to solve the image classification problem. In this paper, we propose a classification model using semantic segmentation to determine the presence or absence of failures such as arcs, disconnections, and cracks in solar panel images obtained using a drone equipped with a thermal imaging camera. In addition, an efficient classification model was implemented by tuning several factors such as data size and type and loss function customization in U-Net, which shows robust classification performance even with a small dataset.

• Journal of Korean Society of Rural Planning
• /
• v.17 no.2
• /
• pp.33-41
• /
• 2011

To evaluate the value of a preserved tree as rural landscape resource, the growth environment and health condition was investigated, and the image evaluation was implemented on land~ape architectural major undergraduate students for zelkova trees in Yesan-gun. The image evaluation results of zelkova trees were as followings; 1) Typical image of preserved tree examined by Semantic Differential Scale were 'Old', 'Big', and 'Good'. 2) The 'big' image of zelkova tree and the height of tree, the width of tree crown, the breast girth of tree, the root girth of tree, the external formation of tree, and the health of tree bark is mutually related. Especially, the correlation between the 'big' and the external formation and the width of tree crown is high. 3) Typical image of preserved tree examined by Likert Scale were 'Natural', 'Green', 'Peaceful', and 'Rural'. 4) The preservation necessity for preserved tree was highly related with the state of ground, and the management necessity for preserved tree was highly related with contamination level and the state of ground. The appropriate management plan for preserved tree are proposed to improve the quality of rural landscape(basis of these results).