• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.12
• /
• pp.4835-4848
• /
• 2020

The recent expansion of the digital content market is increasing the technical demand for various facial image transformations within the virtual environment. The recent image translation technology enables changes between various domains. However, current image-to-image translation techniques do not provide stable performance through unsupervised learning, especially for shape learning in the face transition field. This is because the face is a highly sensitive feature, and the quality of the resulting image is significantly affected, especially if the transitions in the eyes, nose, and mouth are not effectively performed. We herein propose a new unsupervised method that can transform an in-wild face image into another face style through radical transformation. Specifically, the proposed method applies two face-specific feature loss functions for a generative adversarial network. The proposed technique shows that stable domain conversion to other domains is possible while maintaining the image characteristics in the eyes, nose, and mouth.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.10
• /
• pp.35-42
• /
• 2020

In this study, we propose a crack detection method using limited data with a U-Net based image inpainting technique that is a modified unsupervised anomaly detection method. Concrete cracking occurs due to a variety of causes and is a factor that can cause serious damage to the structure in the long term. In general, crack investigation uses an inspector's visual inspection on the concrete surfaces, which is less objective in judgment and has a high possibility of human error. Therefore, a method with objective and accurate image analysis processing is required. In recent years, the methods using deep learning have been studied to detect cracks quickly and accurately. However, when the amount of crack data on the building or infrastructure to be inspected is small, existing crack detection models using it often show a limited performance. Therefore, in this study, an unsupervised anomaly detection method was used to augment the data on the object to be inspected, and as a result of learning using the data, we confirmed the performance of 98.78% of accuracy and 82.67% of harmonic average (F1_Score).

• Journal of the Korea Society of Computer and Information
• /
• v.19 no.8
• /
• pp.11-18
• /
• 2014

In This paper, we propose the unsupervised learning and fuzzy logic-based coupled data classification method base on ART. The unsupervised learning-based data classification helps improve the grouping technique, but decreases the processing efficiency. However, the data classification requires the decision technique to induce high success rate of data classification with optimal threshold. Therefore it is also necessary to solve the uncertainty of the threshold decision. The proposed method deduces the optimal threshold with the designing of fuzzy parameter and rules. In order to evaluate the proposed method, we design the simulation model with the GPCR(G protein coupled receptor) data in cloud computing environment. Simulation results verify the efficiency of our method with the high recognition rate and low processing time.

• The Journal of the Acoustical Society of Korea
• /
• v.40 no.6
• /
• pp.578-586
• /
• 2021

We propose a semi-supervised learning method based on Variational AutoEncoder (VAE) and Unsupervised Data Augmentation (UDA) to improve the performance of an end-to-end speech recognizer. In the proposed method, first, the VAE-based augmentation model and the baseline end-to-end speech recognizer are trained using the original speech data. Then, the baseline end-to-end speech recognizer is trained again using data augmented from the learned augmentation model. Finally, the learned augmentation model and end-to-end speech recognizer are re-learned using the UDA-based semi-supervised learning method. As a result of the computer simulation, the augmentation model is shown to improve the Word Error Rate (WER) of the baseline end-to-end speech recognizer, and further improve its performance by combining it with the UDA-based learning method.

• Korean Journal of Remote Sensing
• /
• v.24 no.4
• /
• pp.341-349
• /
• 2008

This study proposes new automatic thresholding method, which is important step for detecting binary change/non-change information using satellite images. Result value through pixel-based similarity measurement is calculated cumulatively with regular interval, and thresholding is pointed at the steep slope position. The proposed method is assessed in comparison with expectation-maximization algorithm and coner method using synthetic images, ALI images, and Hyperion images. Throughout the results, we validated that our method can guarantee the similar accuracy with previous algorithms. It is simpler than EM algorithm, and can be applied to the binormal histogram unlike the coner method.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.11 no.6
• /
• pp.549-557
• /
• 2001

Morphological analysis and POS tagging require a dictionary for the language at hand . In this fashion though it is impossible to analyze a language a dictionary. We also have difficulty if significant portion of the vocabulary is new or unknown . This paper explores the possibility of learning morphology of an agglutinative language. in particular Korean language, without any prior lexical knowledge of the language. We use unsupervised learning in that there is no instructor to guide the outcome of the learner, nor any tagged corpus. Here are the main characteristics of the approach: First. we use only raw corpus without any tags attached or any dictionary. Second, unlike many heuristics that are theoretically ungrounded, this method is based on statistical methods , which are widely accepted. The method is currently applied only to Korean language but since it is essentially language-neutral it can easily be adapted to other agglutinative languages.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.24 no.10B
• /
• pp.1946-1952
• /
• 1999

A new image compression method to preserve edge characteristics in reconstructed images using an unsupervised learning neural is proposed in this paper. By the unsupervised competitive learning which generalizes previously proposed Centroid Neural Network(CNN) algorithm with the geometric characteristics of edge area and statistical characteristics of image data, more codevectors are allocated in the edge areas to provide the more accurate edges in reconstructed image. Experimental results show that the proposed method gives improved edge in reconstructed images when compared with SOM, Modified SOM and M/R-CNN.

• ETRI Journal
• /
• v.32 no.5
• /
• pp.810-818
• /
• 2010

This paper presents a new framework for integrating untranscribed spoken content into the acoustic training of an automatic speech recognition system. Untranscribed spoken content plays a very important role for under-resourced languages because the production of manually transcribed speech databases still represents a very expensive and time-consuming task. We proposed two new methods as part of the training framework. The first method focuses on combining initial acoustic models using a data-driven metric. The second method proposes an improved acoustic training procedure based on unsupervised transcriptions, in which word endings were modified by broad phonetic classes. The training framework was applied to baseline acoustic models using untranscribed spoken content from parliamentary debates. We include three types of acoustic models in the evaluation: baseline, reference content, and framework content models. The best overall result of 18.02% word error rate was achieved with the third type. This result demonstrates statistically significant improvement over the baseline and reference acoustic models.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.405-405
• /
• 2000

We designed the artificial olfactory sensing system (Electronic Nose) using MOS type sensor array fur recognizing and analyzing odour. The response of individual sensors of sensor array, each processing a slightly different response towards the sample volatiles, can provide enough information to discriminate between sample odours. In this paper, we applied clustering algorithm for dimension reduction, such as linear projection mapping (PCA method), nonlinear mapping (Sammon mapping method) and the combination of PCA and Sammon mapping having a better discriminating ability. The odours used are VOC (Volatile chemical compound) and Toxic gases.

• Proceedings of the Safety Management and Science Conference
• /
• 2008.04a
• /
• pp.185-196
• /
• 2008

The objective of this study is to evaluate the mechanical behaviors and structural integrity of the weldment of high strength steel by using an acoustic emission (AE) techniques. Simple tension and AE tests were conducted against the 3 kind of welding test specimens. In order to analysis the effectiveness of weldability, joinability and structural integrity, we used K-means clustering method as a unsupervised learning pattern recognition algorithm for obtained multivariate AE main data sets, such as AE counts, energy, amplitude, hits, risetime, duration, counts to peak and rms signals. Through the experimental results, the effectiveness of the proposed method is discussed.