• The Journal of the Acoustical Society of Korea
• /
• v.38 no.6
• /
• pp.703-709
• /
• 2019

This paper proposes speech recognition systems employing Convolutional Neural Network (CNN) and Long Short Term Memory (LSTM) structures combined with Hidden Markov Moldel (HMM) to effectively recognize the speech of VeloPharyngeal Insufficiency (VPI) patients, and compares the recognition performance of the systems to the Gaussian Mixture Model (GMM-HMM) and fully-connected Deep Neural Network (DNNHMM) based speech recognition systems. In this paper, the initial model is trained using normal speakers' speech and simulated VPI speech is used for generating a prior model for speaker adaptation. For VPI speaker adaptation, selected layers are trained in the CNN-HMM based model, and dropout regulatory technique is applied in the LSTM-HMM based model, showing 3.68 % improvement in recognition accuracy. The experimental results demonstrate that the proposed LSTM-HMM-based speech recognition system is effective for VPI speech with small-sized speech data, compared to conventional GMM-HMM and fully-connected DNN-HMM system.

• The Journal of the Acoustical Society of Korea
• /
• v.26 no.1
• /
• pp.25-31
• /
• 2007

This paper describes an attempt to prevent the external acoustic noise from being misrecognized as the speech recognition target. For this, in the speech activity detection process for the speech recognition, it confirmed besides the acoustic energy to the lip movement image signal of a speaker. First of all, the successive images are obtained through the image camera for PC. The lip movement whether or not is discriminated. And the lip movement image signal data is stored in the shared memory and shares with the recognition process. In the meantime, in the speech activity detection Process which is the preprocess phase of the speech recognition. by conforming data stored in the shared memory the acoustic energy whether or not by the speech of a speaker is verified. The speech recognition processor and the image processor were connected and was experimented successfully. Then, it confirmed to be normal progression to the output of the speech recognition result if faced the image camera and spoke. On the other hand. it confirmed not to output of the speech recognition result if did not face the image camera and spoke. That is, if the lip movement image is not identified although the acoustic energy is inputted. it regards as the acoustic noise.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.10
• /
• pp.217-223
• /
• 2009

Vocabulary recognition system of Mobile terminal is executed statistical method for vocabulary recognition and used statistical grammar recognition system using N-gram. If limit arithmetic processing capacity in memory of vocabulary to grow then vocabulary recognition algorithm complicated and need a large scale search space and many processing time on account of impossible to process. This study suggest vocabulary recognition optimize using MLHF System. MLHF separate acoustic search and lexical search system using FLaVoR. Acoustic search feature vector of speech signal extract using HMM, lexical search recognition execution using Levenshtein distance algorithm. System performance as a result of represent vocabulary dependence recognition rate of 98.63%, vocabulary independence recognition rate of 97.91%, represent recognition speed of 1.61 second.

• MALSORI
• /
• no.50
• /
• pp.51-69
• /
• 2004

In this paper, we present a one-pass semi-dynamic network decoding framework that inherits both advantages of fast decoding speed from static network decoders and memory efficiency from dynamic network decoders. Our method is based on the novel language model network representation that is essentially of finite state machine (FSM). The static network derived from the language model network [1][2] is partitioned into smaller subnetworks which are static by nature or self-structured. The whole network is dynamically managed so that those subnetworks required for decoding are cached in memory. The network is near-minimized by applying the tail-sharing algorithm. Our decoder is evaluated on the 25k-word Korean broadcast news transcription task. In case of the search network itself, the network is reduced by 73.4% from the tail-sharing algorithm. Compared with the equivalent static network decoder, the semi-dynamic network decoder has increased at most 6% in decoding time while it can be flexibly adapted to the various memory configurations, giving the minimal usage of 37.6% of the complete network size.

• ETRI Journal
• /
• v.44 no.4
• /
• pp.672-685
• /
• 2022

This article proposes a subject-independent application of brain-computer interfacing (BCI). A 32-channel Electroencephalography (EEG) device is used to measure imagined speech (SI) of four words (sos, stop, medicine, washroom) and one phrase (come-here) across 13 subjects. A deep long short-term memory (LSTM) network has been adopted to recognize the above signals in seven EEG frequency bands individually in nine major regions of the brain. The results show a maximum accuracy of 73.56% and a network prediction time (NPT) of 0.14 s which are superior to other state-of-the-art techniques in the literature. Our analysis reveals that the alpha band can recognize SI better than other EEG frequencies. To reinforce our findings, the above work has been compared by models based on the gated recurrent unit (GRU), convolutional neural network (CNN), and six conventional classifiers. The results show that the LSTM model has 46.86% more average accuracy in the alpha band and 74.54% less average NPT than CNN. The maximum accuracy of GRU was 8.34% less than the LSTM network. Deep networks performed better than traditional classifiers.

• The Korean Journal of Legal Medicine
• /
• v.39 no.4
• /
• pp.115-119
• /
• 2015

We investigated event-related potentials (ERPs) to estimate the accuracy of eyewitness memories. Participants watched videos of vehicles being driven dangerously, from an anti-impaired driving initiative. The four-letter license plates of the vehicles were the target stimuli. Random numbers were presented while participants attempted to identify the license plate letters, and electroencephalograms were recorded. There was a significant difference in activity 300-500 milliseconds after stimulus onset, between target stimuli and random numbers. This finding contributes to establishing an eyewitness recognition model where different ERP components may reflect more explicit memory that is dissociable from recollection.

• Journal of Information Technology Services
• /
• v.18 no.1
• /
• pp.1-12
• /
• 2019

The purpose of this study is to find out what type of skip button used in forced exposure advertising is the most positive to the users. The four types of skip buttons were produced for the experiment and tested by survey and eye tracker to reveal the effects of the skip buttons on perceived intrusion, advertising attention, attitude toward advertising, and memory consisting of recall and recognition. Out of 80 participants, 20 were randomly assigned to the specific type of skip button group. The results showed that there is no statistical difference in advertising attention, perceived intrusiveness and attitude toward advertising. However, the recall and recognition rate are the highest in the static text type and kinetic text, product image, and default follow statistically. This study has implications for using skip buttons as a major variable for inventory of TrueView advertising effects and suggests that the amount of information in the image is critical processed by users within very short time.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.11
• /
• pp.5580-5593
• /
• 2019

A malicious Uniform Resource Locator (URL) recognition and detection method based on the combination of Attention mechanism with Convolutional Neural Network and Long Short-Term Memory Network (Attention-Based CNN-LSTM), is proposed. Firstly, the WHOIS check method is used to extract and filter features, including the URL texture information, the URL string statistical information of attributes and the WHOIS information, and the features are subsequently encoded and pre-processed followed by inputting them to the constructed Convolutional Neural Network (CNN) convolution layer to extract local features. Secondly, in accordance with the weights from the Attention mechanism, the generated local features are input into the Long-Short Term Memory (LSTM) model, and subsequently pooled to calculate the global features of the URLs. Finally, the URLs are detected and classified by the SoftMax function using global features. The results demonstrate that compared with the existing methods, the Attention-based CNN-LSTM mechanism has higher accuracy for malicious URL detection.

• International Journal of Contents
• /
• v.15 no.4
• /
• pp.8-15
• /
• 2019

This paper presents a solution of the 'Quick, Draw! Doodle Recognition Challenge' hosted by Google. Doodles are drawings comprised of concrete representational meaning or abstract lines creatively expressed by individuals. In this challenge, a doodle is presented as a sequence of sketches. From the view of at the sketch level, to learn the pattern of strokes representing a doodle, we propose a sequential model stacked with multiple convolution layers and Long Short-Term Memory (LSTM) cells following the attention mechanism [15]. From the view at the image level, we use multiple models pre-trained on ImageNet to recognize the doodle. Finally, an ensemble and a post-processing method using the minimum cost flow algorithm are introduced to combine multiple models in achieving better results. In this challenge, our solutions garnered 11th place among 1,316 teams. Our performance was 0.95037 MAP@3, only 0.4% lower than the winner. It demonstrates that our method is very competitive. The source code for this competition is published at: https://github.com/ngxbac/Kaggle-QuickDraw.

• ETRI Journal
• /
• v.42 no.5
• /
• pp.761-772
• /
• 2020

This paper presents an automatic proficiency assessment method for a non-native Korean read utterance using bidirectional long short-term memory (BLSTM)-based acoustic models (AMs) and speech data augmentation techniques. Specifically, the proposed method considers two scenarios, with and without prompted text. The proposed method with the prompted text performs (a) a speech feature extraction step, (b) a forced-alignment step using a native AM and non-native AM, and (c) a linear regression-based proficiency scoring step for the five proficiency scores. Meanwhile, the proposed method without the prompted text additionally performs Korean speech recognition and a subword un-segmentation for the missing text. The experimental results indicate that the proposed method with prompted text improves the performance for all scores when compared to a method employing conventional AMs. In addition, the proposed method without the prompted text has a fluency score performance comparable to that of the method with prompted text.