• Proceedings of the KIEE Conference
• /
• 2006.07d
• /
• pp.2083-2084
• /
• 2006

본 논문에서는 조명에 변화에 강인하고 기존의 퍼지 색상 필터보다 정확하고 빠른 얼굴 감지 알고리즘 이용하여 얼굴을 인식하고 얼굴로부터 특징점(눈, 눈썹, 입)틀을 추출하고 추출된 특징점을 이용하여 감성을 판별하는 방법을 제안한다. 향상된 얼굴 인식 기술이란 퍼지 색상 필터의 단점이 영상의 크기와 성능에 따라 처리속도가 느려지는 것을 보완하기 위하여 최소한의 규칙을 사용하여 얼굴 후보 영역을 선별 적용하여 얼굴영역을 추출하는 기법을 말한다. 이렇게 추출된 얼굴영역에서 감정이 변화 할 때 가장 두드러지게 변화를 나타내는 눈, 눈썹 그리고 입의 특징점을 이용하여 감성을 분류한다.

• Proceedings of the KIEE Conference
• /
• 2006.10c
• /
• pp.214-216
• /
• 2006

Expressive face and human body gestures are among the main non-verbal communication channels in human-human interaction. Understanding human emotions through body gesture is one of the necessary skills both for humans and also for the computers to interact with their human counterparts. Gesture analysis is consisted of several processes such as detecting of hand, extracting feature, and recognizing emotions. Skin color information for tracking hand gesture is obtained from face detection region. We have revealed relationships between paricular body movements and specific emotions by using HMM(Hidden Markov Model) classifier. Performance evaluation of emotional human body recognition has experimented.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2015.07a
• /
• pp.421-424
• /
• 2015

컴퓨터 기술이 발전하고 컴퓨터 사용이 일반화 되면서 휴먼 인터페이스에 대한 많은 연구들이 진행되어 왔다. 휴먼 인터페이스에서 감정을 인식하는 기술은 컴퓨터와 사람간의 상호작용을 위해 중요한 기술이다. 감정을 인식하는 기술에서 분류 정확도를 높이기 위해 특징벡터를 정확하게 추출하는 것이 중요하다. 본 논문에서는 정확한 피치검출을 위하여 음성신호에서 음성 구간과 비 음성구간을 추출하였으며, Speech Processing 분야에서 사용되는 전 처리 기법인 저역 필터와 유성음 추출 기법, 후처리 기법인 Smoothing 기법을 사용하여 피치 검출을 수행하고 비교하였다. 그 결과, 전 처리 기법인 유성음 추출 기법과 후처리 기법인 Smoothing 기법은 피치 검출의 정확도를 높였고, 저역 필터를 사용한 경우는 피치 검출의 정확도가 떨어트렸다.

• Annual Conference of KIPS
• /
• 2022.11a
• /
• pp.580-582
• /
• 2022

본 본문은 영상을 통해 반려동물의 행동과 감정을 인식하는 연구이다. 이러한 연구는 반려동물과 인간의 삶을 향상시키는데 도움을 줄 수 있다. 제안된 시스템의 유용성은 실험을 통해 검증하였고, 어느 정도 의미 있는 결과를 얻을 수 있었다.

• Annual Conference of KIPS
• /
• 2022.11a
• /
• pp.818-820
• /
• 2022

The keyword 'healing' is essential to the competitive society and culture of Koreans. In addition, as the time at home increases due to COVID-19, the demand for indoor healing services has increased. Therefore, this thesis analyzes the user's facial expression so that people can receive various 'customized' healing services indoors, and based on this, provides lighting, ASMR, video recommendation service, and facial expression recording service.The user's expression was analyzed by applying the ensemble algorithm to the expression prediction results of various CNN models after extracting only the face through object detection from the image taken by the user.

• The KIPS Transactions:PartB
• /
• v.17B no.4
• /
• pp.263-274
• /
• 2010

To provide context-aware music recommendation service, first of all, we need to catch music mood that a user prefers depending on his situation or context. Among various music characteristics, music mood has a close relation with people‘s emotion. Based on this relationship, some researchers have studied on music mood detection, where they manually select a representative segment of music and classify its mood. Although such approaches show good performance on music mood classification, it's difficult to apply them to new music due to the manual intervention. Moreover, it is more difficult to detect music mood because the mood usually varies with time. To cope with these problems, this paper presents an automatic method to classify the music mood. First, a whole music is segmented into several groups that have similar characteristics by structural information. Then, the mood of each segments is detected, where each individual's preference on mood is modelled by regression based on Thayer's two-dimensional mood model. Experimental results show that the proposed method achieves 80% or higher accuracy.

• Science of Emotion and Sensibility
• /
• v.15 no.2
• /
• pp.259-268
• /
• 2012

In a P300-based concealed information test (P300 CIT), the result of the test is greatly affected by the value of the probe stimulus. With a probe stimulus of low value, the detection rate decreases. The aim of this study was to determine whether the pupil-based concealed information test (Pupil CIT) could be used in addition to the P300 CIT for the probes of low value. Participants were told to choose one card from a deck of five cards (space 2, 3, 4, 5, 6), Then a P300 CIT and a Pupil CIT for the selected card were administered. P300s were measured at 3 scalp sites (Fz, Cz, and Pz), and the pupil sizes of left and right eyes were recorded. The P300 amplitude measured at Fz, Cz, and Pz was significantly different between the probe and irrelevant stimuli. And, in the Pupil CIT, the pupil size was also different between the two stimuli for both eyes. The detection rates of the P300 CIT were 44% at Fz and Cz sites and 36% at Pz site. And the detection rates of the Pupil CIT were 52% for the left eye and 60% for the right eye. There is a trend that the detection rate of the Pupil CIT was higher than that of the P300 CIT, but the difference didn't reach significance partly because of the relatively small sample size. The correlation between the decision based on the P300 CIT and that based on the Pupil CIT was not significant. As a conclusion, it is recommended to use a Pupil CIT instead of a P300 CIT when the value of the probe is low. And a combination of the measures may be superior to either one of them in detection rate.

• Science of Emotion and Sensibility
• /
• v.11 no.2
• /
• pp.235-244
• /
• 2008

This study is to develop automatic extraction system of continuous blood pressure using ECG (Electrocardiogram) and PPG(Photoplethysmography) for u-health care technology. PTT (Pulse Transit Time) was determined from peak difference between ECG and PPG and its inverse made to get blood pressure. Since the peaks were vulnerable to be contaminated from noise and variation of amplitude, this study developed the adaptive algorithm for peak calculation in any noise condition. The developed method of the adaptive peak calculation was proven to make the standard deviations of PPT decrease to 28% and the detection of noise increase to 18%. Also, the correlation model such as blood pressure = -0.044 $\cdot$ PTT + 133.592 has successfully been determined for predicting the continuous pressure measured without using cuff but with using PPG and ECG, only.

• Science of Emotion and Sensibility
• /
• v.11 no.2
• /
• pp.207-214
• /
• 2008

The inspection processing is for reducing loss which occurs fault because of fabric appearance. Up to now inspection machine which is used from inspection process is classified with the macrography inspection machine and the full automatic inspection machine. The macrography inspection machine is low price and efficient equipment but does not record information of fault. On the other side, the automatic inspection machine is high price, also the detection rate of one changes with effect of environment variable but able to record information of fault. It developed semi-automatic cloth inspection machine with the weak point of the macrography inspection machine and the automatic inspection machine was complemented. And when it uses information which was collected by semi-automatic cloth inspection machine, the loss rate of original fabric is able to calculate. So sewing factories will be able to predict fabric consuming quantity.

• Journal of the Korea Convergence Society
• /
• v.9 no.7
• /
• pp.25-31
• /
• 2018

Online news has replaced the traditional newspaper and has brought about a profound transformation in the way we access and share information. News websites have had the ability for users to post comments for quite some time, and some have also begun to crowdsource reactions to news articles. The field of sentiment analysis seeks to computationally model the emotions and reactions experienced when presented with text. In this work, we analyze more than 100,000 news articles over ten categories with five user-generated emotional annotations to determine whether or not these reactions have a mathematical correlation to the news body text and propose a simple sentiment analysis algorithm that requires minimal preprocessing and no machine learning. We show that it is effective even for a morphologically complex language like Korean.