• Phonetics and Speech Sciences
• /
• v.12 no.4
• /
• pp.63-71
• /
• 2020

Speech recognition is one of the areas actively commercialized using deep learning and machine learning techniques. However, the majority of speech recognition systems on the market are developed on data with limited diversity of speakers and tend to perform well on typical adult speakers only. This is because most of the speech recognition models are generally learned using a speech database obtained from adult males and females. This tends to cause problems in recognizing the speech of the elderly, children and people with dialects well. To solve these problems, it may be necessary to retain big database or to collect a data for applying a speaker adaptation. However, this paper proposes that a new end-to-end speech recognition method consists of an acoustic augmented recurrent encoder and a transformer decoder with linguistic prediction. The proposed method can bring about the reliable performance of acoustic and language models in limited data conditions. The proposed method was evaluated to recognize Korean elderly and children speech with limited amount of training data and showed the better performance compared of a conventional method.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.8
• /
• pp.365-370
• /
• 2023

This paper aims to compare the performance of speech data classification into two groups, adult and elderly, based on the acoustic and linguistic characteristics that change due to aging, such as changes in respiratory patterns, phonation, pitch, frequency, and language expression ability. For acoustic features we used attributes related to the frequency, amplitude, and spectrum of speech voices. As for linguistic features, we extracted hidden state vector representations containing contextual information from the transcription of speech utterances using KoBERT, a Korean pre-trained language model that has shown excellent performance in natural language processing tasks. The classification performance of each model trained based on acoustic and linguistic features was evaluated, and the F1 scores of each model for the two classes, adult and elderly, were examined after address the class imbalance problem by down-sampling. The experimental results showed that using linguistic features provided better performance for classifying adult and elderly than using acoustic features, and even when the class proportions were equal, the classification performance for adult was higher than that for elderly.

• The Korean Journal of Applied Statistics
• /
• v.36 no.1
• /
• pp.33-48
• /
• 2023

In this paper we consider automatic speech recognition (ASR) for Korean speech data in which elderly persons randomly speak a sequence of words such as animals and vegetables for one minute. Most of the speakers are over 60 years old and some of them are dementia patients. The goal is to compare deep-learning based ASR models for such data and to find models with good performance. ASR is a technology that can recognize spoken words and convert them into written text by computers. Recently, many deep-learning models with good performance have been developed for ASR. Training data for such models are mostly composed of the form of sentences. Furthermore, the speakers in the data should be able to pronounce accurately in most cases. However, in our data, most of the speakers are over the age of 60 and often have incorrect pronunciation. Also, it is Korean speech data in which speakers randomly say series of words, not sentences, for one minute. Therefore, pre-trained models based on typical training data may not be suitable for our data, and hence we train deep-learning based ASR models from scratch using our data. We also apply some data augmentation methods due to small data size.

• International Journal of Contents
• /
• v.8 no.3
• /
• pp.26-33
• /
• 2012

The animal naming test (ANT) is known to be influenced not only by age, gender, and education but only by ethnicity, culture, and language. Thus, population-specific norm considering these variables needs to be developed for Korean-speaking elderly. We evaluated 185 healthy elderly people with five measures. Education was the single statistically independent correlate of the total number of words ($R^2$ = .312, p = .038). After adjusting for education, there was slightly significant negative correlation (r = -.215, p = .049) between age and total number of words. Mean number of words produced was $13.71{\pm}3.09$. The production frequency was negatively correlated with the typicality rating (r = -0.41, p < .05). The concrete and exact scoring rule could be set up in the comparison of naming performance between a normal and patient with neuro-linguistic disorder and its data could be utilized in a differential diagnosis for patients with neurological disorders.

• Journal of Korean society of Dental Hygiene
• /
• v.19 no.4
• /
• pp.577-589
• /
• 2019

Objectives: The purpose of this study was to investigate the effects of general and oral health on quality of life in the elderly living alone and with family. Methods: We analyzed data from the $6^{th}$ Korea National Health and Nutrition Examination Survey. Distribution of the elderly living alone and with family based on the general characteristics and general and oral health was analyzed using complex-sample chi-square tests. Multiple logistic regression was used to analyze the factors affecting quality of life by calculating the 95% confidence intervals. Results: In the elderly living alone, the quality of life significantly correlated with restriction of activity, perceived general and oral health status, perceived stress, and speech difficulties. Further, in the elderly living with family, lower quality of life significantly correlated with restriction of activity, perceived health status, walking days per week, life time smoking history, Community Periodontal Index, and chewing and speech difficulties. Conclusions: The elderly are concerned with self-maintenance of general and oral health. Therefore, systematic policies related to health services need to be developed and operated at the national level. It is especially necessary to take social interest in the elderly living alone and a more continuous and professional approach in their health care.

• 한국노년학
• /
• v.41 no.4
• /
• pp.505-525
• /
• 2021

Purpose. The purpose of this study is to examine how the phrase "old person" are expressed and used in the online sphere. Based on the theoretical concept of stigma, this study investigates the images and attitudes in society toward the elderly, and the characteristics of hate speech aimed at the elderly. Method. This study conducted text mining based on social big data using anonymous conversations. Results. It was confirmed that the elderly images shared online were generally negative. The attitudes expressed toward them also tended to be negative due to the negative images that are propagated of the elderly. The hate speech relating to the elderly, in usages such as 'Teul-ttag' and 'Kon-dae', were mainly identified in comments that negatively evaluate the elderly, and these expressions demonstrate the depth of hate and discrimination towards the elderly who are considered burdensome by young people. Interestingly, the hateful expressions towards the elderly were found more with regard to issues related to politics and economics and not just any content about the elderly. Conclusions. This study discussed the ways and means to enhance inter-generational understanding and solidity.

• 재활복지
• /
• v.21 no.1
• /
• pp.193-212
• /
• 2017

The elderly often produce irrelevant speech and get off-topic more easily than the young; the former also has difficulty generating fewer syntactic structures and makes errors of grammatical morphemes. In particular, the elderly might have more difficulty writing since it requires more complex cognitive processes than storytelling. The participants in this study were 32 young people and 32 older people. They were asked to write a short story of Korean fairy tale('Heungbu Nolbu'). The data was analyzed in narrative composition and syntactic structures. The study revealed the following: First, in composition aspects, the elderly group showed significantly lower total number of story grammar and episodes. In addition, the elderly produced more off topic statements. Second, in syntactic aspects, although there was no significant difference in the number of producing complex sentences between two groups, the elderly group generated more inadequate cohesive devices and used fewer relative and adverbial clauses. These findings suggest that the elderly have a tendency to perform tasks by producing more off-topic statements and shows decreasing coherence by using lower number of relative and adverbial clauses. However, this study also uncovers that the elderly were able to write more complex and longer sentences using visual feedback.

• Phonetics and Speech Sciences
• /
• v.4 no.4
• /
• pp.155-162
• /
• 2012

The population of Korea is ageing as the number of elderly people increases due to improvements in health care and diet. Accordingly, it is expected that interest in how to live actively during the years after retirement and how to communicate effectively will increase the demand for voice improvement methods and technology. However, the criteria to evaluate the voice strength and characteristics of the elderly are lacking. In this study, we analyzed the acoustic characteristics of elderly women living in the community according to residential status and mental health status (e.g. depressive mood). Accordingly, we selected women (n=63) above the age of 65 age who were living in the Seoul metropolitan area and Daegu Gyeongbuk. The selected subjects were divided into two groups: a normal speaker group (n=40) and a speaker group comprised of those suffering from depressive mood (n=23). This study analyzed the voice characteristics of subjects based on collected data through the sustained phonation of the vowel /a/. It was shown that there were differences among MPT, F0, Jitter, Shimmer and NHR depending on location of residence but no difference with regard to depressive mood. Therefore, we must consider location of residence in elderly as the key factor in demonstrating the voice norms of seniors.

• Speech Sciences
• /
• v.15 no.3
• /
• pp.95-102
• /
• 2008

Multiple physical changes of the larynx and its components occur with age. Vocal pitch, commonly expressed through measures of fundamental frequency (Fo) relate to physical conditions of the larynx. Available data is lacking for the senescent voice, and should be applied to the of changes of elderly speakers' Fo characteristics. The purpose of this study was to investigate the Fo of normal elderly speaker's voice. A total of 406 normal elderly speakers (207 males and 199 females) participated in this experiment. Age ranged from 60 years to 89 years. The subjects were asked to produce sustained corner vowels (/a/ /i/ /u/) three times each and the data were analyzed using the MDVP of CSL. According to the results of this study, the mean Fo from the ages of 60's to 80's shows 143.95Hz(SD 13.94) for men and 185.42Hz (SD 15.29) for women. For men, a significant change is found as a function of age in the Fo (F=16.181, p<.05). A post-hoc Scheffe test revealed significant differences between the Fo data of subjects aged 60's and 70's, 60's and 80's. For women, a significant change is found as a function of age in the Fo (F=49.013, p<.05). A post-hoc $Scheff'{e}$ test revealed significant differences between the Fo data of subjects in their 60's and 70's, 70's and 80's, 60's and 80's. The Fo of men goes up from their 60's to 80's gradually, whereas the Fo of women goes down gradually until their 70's, and after their 70's it again increases. It has been known that diminishing estrogen levels in women in old age may be a factor in lowering Fo, whereas diminishing testosterone levels in men may contribute to a rising Fo. This result may be used as some meaningful guideline and lead the basic data to differentiate between normal aged voice and aged voice disorders.

• Journal of Korean Academy of Dental Administration
• /
• v.8 no.1
• /
• pp.30-36
• /
• 2020

This study is conducted on 1,725 elderly people over 65 years of age using 2018 data obtained from the 7th National Health and Nutrition Survey (KNHANES) data. In this study, an analysis is performed considering the general characteristics of the elderly and their oral health status (authoring discomfort, speech problems, etc.) to confirm the relationship between the elderly's unmet dental experience and depressive experience. The results of this study showed that depressive experiences by the elderly resulted in unmet dental medical experiences, and it was also found that the income level and the complaint of chewing discomfort had an effect. Based on these results, it is believed that oral health policies should be developed to improve the unmet dental medical experience by considering the socio-economic level of the elderly and depressive experiences. This policy development is expected to lead not only to the improvement of oral health for the elderly, but also to improve the quality of life for the elderly through health promotion.