• The Journal of the Acoustical Society of Korea
• /
• v.20 no.2E
• /
• pp.17-24
• /
• 2001

In this paper, we propose the bootstrap and aggregating (bagging) vector quantization (VQ) classifier to improve the performance of the text-independent speaker recognition system. This method generates multiple training data sets by resampling the original training data set, constructs the corresponding VQ classifiers, and then integrates the multiple VQ classifiers into a single classifier by voting. The bagging method has been proven to greatly improve the performance of unstable classifiers. Through two different experiments, this paper shows that the VQ classifier is unstable. In one of these experiments, the bias and variance of a VQ classifier are computed with a waveform database. The variance of the VQ classifier is compared with that of the classification and regression tree (CART) classifier[1]. The variance of the VQ classifier is shown to be as large as that of the CART classifier. The other experiment involves speaker recognition. The speaker recognition rates vary significantly by the minor changes in the training data set. The speaker recognition experiments involving a closed set, text-independent and speaker identification are performed with the TIMIT database to compare the performance of the bagging VQ classifier with that of the conventional VQ classifier. The bagging VQ classifier yields improved performance over the conventional VQ classifier. It also outperforms the conventional VQ classifier in small training data set problems.

• Journal of Broadcast Engineering
• /
• v.28 no.2
• /
• pp.178-184
• /
• 2023

Finding texts in general scene images and recognizing their contents is a very important task that can be used as a basis for robot vision, visual assistance, and so on. However, for the low-resolution text images, the degradations, such as noise or blur included in text images, are more noticeable, which leads to severe performance degradation of text recognition accuracy. In this paper, we propose a new Korean text image super-resolution based on a Transformer-based model, which generally shows higher performance than convolutional neural networks. In the experiments, we show that text recognition accuracy for Korean text images can be improved when our proposed text image super-resolution method is used. We also propose a new Korean text image dataset for training our model, which contains massive HR-LR Korean text image pairs.

• Journal of the Korea Society of Computer and Information
• /
• v.22 no.9
• /
• pp.17-24
• /
• 2017

Core citation recognition is to identify influential ones among the prior articles that a scholarly article cite. Previous approaches have employed citing-text occurrence information, textual similarities between citing and cited article, etc. This study proposes a term-based approach to core citation recognition, which exploits the importance of individual terms appearing in in-text citation to calculate influence-strength for each cited article. Term importance is computed using various frequency information such as term frequency(tf) in in-text citation, tf in the citing article, inverse sentence frequency in the citing article, inverse document frequency in a collection of articles. Experiments using a previous test set consisting of computational linguistics articles show that the term-based approach performs comparably with the previous approaches. The proposed technique could be easily extended by employing other term units such as n-grams and phrases, or by using new term-importance formulae.

• ETRI Journal
• /
• v.46 no.1
• /
• pp.96-105
• /
• 2024

We propose a method to automatically predict Alzheimer's disease from speech data using the ChatGPT large language model. Alzheimer's disease patients often exhibit distinctive characteristics when describing images, such as difficulties in recalling words, grammar errors, repetitive language, and incoherent narratives. For prediction, we initially employ a speech recognition system to transcribe participants' speech into text. We then gather opinions by inputting the transcribed text into ChatGPT as well as a prompt designed to solicit fluency evaluations. Subsequently, we extract embeddings from the speech, text, and opinions by the pretrained models. Finally, we use a classifier consisting of transformer blocks and linear layers to identify participants with this type of dementia. Experiments are conducted using the extensively used ADReSSo dataset. The results yield a maximum accuracy of 87.3% when speech, text, and opinions are used in conjunction. This finding suggests the potential of leveraging evaluation feedback from language models to address challenges in Alzheimer's disease recognition.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.7
• /
• pp.1713-1723
• /
• 1999

Optical character recognition(OCR) might be the most plausible method in building database out of printed matters. This paper describes the points to be considered when one selects an OCR system in order to build database. Based on the considerations, we evaluated four commercial OCR systems, and chose one which shows the best recognition rate to build OCT-text database. The subject text, the KT-test collection, is a set of abstracts from proceedings of different printing quality, fonts, and formats. KT-test collection is also provided with typed text database. Recognition rate was calculated by comparing the recognition result with the typed text. No preprocessing such as learning and slant correction was applied to the recognition process in order to simulate a practical environment. The result shows 90.5% of character recognition rate over 970 abstracts. This recognition rate is still insufficient for practical use. The errors in OCR texts are different from those of manually typed texts. In this paper, we classify the errors in OCR texts for the further research.

• The Journal of the Korea Contents Association
• /
• v.5 no.2
• /
• pp.229-242
• /
• 2005

We constructed a full-text retrieval system for the scanned Hangul document images. The system consists of three parts; preprocessing, recognition, and retrieval components. The retrieval algorithm uses recognition results up to k-ranks. The algorithm is not only insensitive to the recognition errors, but also has the advantage of user-controllable recall and precision. For the objective performance evaluation, we used the scanned images of the Journal of Korea Information Science Society provided by KISTI. The system was shown to be practical through theevaluationofrecognitionandretrievalrates.

• International Journal of Knowledge Content Development & Technology
• /
• v.7 no.1
• /
• pp.101-120
• /
• 2017

Healthcare professionals have edited many health questions (HQs) and their answers for healthcare consumers on the Internet. The HQs provide both readable and reliable health information, and hence retrieval of those HQs that are relevant to a given question is essential for health education and promotion through the Internet. However, retrieval of relevant HQs needs to be based on the recognition of the intention of each HQ, which is difficult to be done by predefining syntactic and semantic rules. We thus model the intention recognition problem as a text classification problem, and develop two techniques to improve a learning-based text classifier for the problem. The two techniques improve the classifier by location-based and area-based feature weightings, respectively. Experimental results show that, the two techniques can work together to significantly improve a Support Vector Machine classifier in both the recognition of HQ intentions and the retrieval of relevant HQs.

• Smart Media Journal
• /
• v.11 no.8
• /
• pp.65-72
• /
• 2022

Due to COVID-19, the importance of non-face-to-face counseling is increasing as the face-to-face counseling method has progressed to non-face-to-face counseling. The advantage of non-face-to-face counseling is that it can be consulted online anytime, anywhere and is safe from COVID-19. However, it is difficult to understand the client's mind because it is difficult to communicate with non-verbal expressions. Therefore, it is important to recognize emotions by accurately analyzing text and voice in order to understand the client's mind well during non-face-to-face counseling. Therefore, in this paper, text data is vectorized using FastText after separating consonants, and voice data is vectorized by extracting features using Log Mel Spectrogram and MFCC respectively. We propose a multi-emotion recognition model that recognizes five emotions using vectorized data using an LSTM model. Multi-emotion recognition is calculated using RMSE. As a result of the experiment, the RMSE of the proposed model was 0.2174, which was the lowest error compared to the model using text and voice data, respectively.

• Proceedings of the IEEK Conference
• /
• 2001.06c
• /
• pp.5-8
• /
• 2001

The computer programs based upon graphic user interface(GUI) became commonplace with the advance of computer technology. Nevertheless, programs for the visually-handicapped have still remained at the level of TTS(text to speech) programs and this prevents many visually-handicapped from enjoying the pleasure and convenience of the information age. This paper is, paying attention to the importance of character recognition in images, about the configuration of the system that converts text in the image selected by a user to the speech by extracting the character part, and carrying out character recognition.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1994.06a
• /
• pp.1082-1087
• /
• 1994

A set of text database is indispensable to the probabilistic models for speech recognition, linguistic model, and machine translation. We introduce an environment to canstruct text databases : an automatic tagging system and a set of tools for lexical knowledge acquisition, which provides the facilities of automatic part of speech recognition and guessing.