• ETRI Journal
• /
• v.46 no.1
• /
• pp.71-81
• /
• 2024

Recent advances in deep learning for speech and visual recognition have accelerated the development of multimodal speech recognition, yielding many innovative results. We introduce a Korean audiovisual speech recognition corpus. This dataset comprises approximately 150 h of manually transcribed and annotated audiovisual data supplemented with additional 2000 h of untranscribed videos collected from YouTube under the Creative Commons License. The dataset is intended to be freely accessible for unrestricted research purposes. Along with the corpus, we propose an open-source framework for automatic speech recognition (ASR) and audiovisual speech recognition (AVSR). We validate the effectiveness of the corpus with evaluations using state-of-the-art ASR and AVSR techniques, capitalizing on both pretrained models and fine-tuning processes. After fine-tuning, ASR and AVSR achieve character error rates of 11.1% and 18.9%, respectively. This error difference highlights the need for improvement in AVSR techniques. We expect that our corpus will be an instrumental resource to support improvements in AVSR.

• The Journal of the Acoustical Society of Korea
• /
• v.17 no.4E
• /
• pp.11-16
• /
• 1998

For large vocabulary speech recognition using HMMs, context-dependent subword units have been often employed. However, when context-dependent phone models are used, they result in a system which has too may parameters to train. The problem of too many parameters and too little training data is absolutely crucial in the design of a statistical speech recognizer. Furthermore, when building large vocabulary speech recognition systems, unseen triphone problem is unavoidable. In this paper, we propose the modified phonetic decision tree algorithm for the automatic prediction of unseen triphones which has advantages solving these problems through following two experiments in Japanese contexts. The baseline experimental results show that the modified tree based clustering algorithm is effective for clustering and reducing the number of states without any degradation in performance. The task experimental results show that our proposed algorithm also has the advantage of providing a automatic prediction of unseen triphones.

• Phonetics and Speech Sciences
• /
• v.8 no.2
• /
• pp.23-30
• /
• 2016

This paper describes some implementation schemes of CNN in view of mini-batch DNN training for efficient second order optimization. This uses same procedure updating parameters of DNN to train parameters of CNN by simply arranging an input image as a sequence of local patches, which is actually equivalent with mini-batch DNN training. Through this conversion, second order optimization providing higher performance can be simply conducted to train the parameters of CNN. In both results of image recognition on MNIST DB and syllable automatic speech recognition, our proposed scheme for CNN implementation shows better performance than one based on DNN.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2008.05a
• /
• pp.784-788
• /
• 2008

In this research, we propose a remote speech input unit, a new method of user-friendly speech input in speech recognition system. We focused the user friendliness on hands-free and microphone independence in speech recognition applications. Our module adopts two algorithms, the automatic speech detection and speech enhancement based on the microphone array-based beamforming method. In the performance evaluation of speech detection, within-200msec accuracy with respect to the manually detected positions is about 97percent under the noise environments of 25dB of the SNR. The microphone array-based speech enhancement using the delay-and-sum beamforming algorithm shows about 6dB of maximum SNR gain over a single microphone and more than 12% of error reduction rate in speech recognition.

• Phonetics and Speech Sciences
• /
• v.4 no.3
• /
• pp.95-101
• /
• 2012

In this paper, we introduce the automatic speaker identification system 'SPO(Supreme Prosecutors Office) Verifier'. SPO Verifier is a GMM(Gaussian mixture model)-UBM(universal background model) based automatic speaker recognition system and has been developed using Korean speakers' utterances. This system uses a channel compensation algorithm to compensate recording device characteristics. The system can give the users the ability to manage reference models with utterances from various environments to get more accurate recognition results. To evaluate the performance of SPO Verifier on Korean speakers, we compared this system with one of the most widely used commercial systems in the forensic field. The results showed that SPO Verifier shows lower EER(equal error rate) than that of the commercial system.

• Proceedings of the IEEK Conference
• /
• 2001.09a
• /
• pp.503-506
• /
• 2001

In this paper, we describe the implementation of a microphone input level control algorithm and the speech improvement with this level controller in personal computer environment. The volume of speech obtained through a microphone affects the speech recognition rate directly. Therefore, proper input volume level control is desired fur better recognition. We considered some conditions for the successful volume controller implementation firstly, then checked its usefulness on our speech recognition system with common office environment speech database. Cepstral mean subtraction is also utilized far the channel-effect compensation of the database. Our implemented controller achieved approximately 50% reduction, i.e., improvement in speech recognition error rate.

• MALSORI
• /
• no.46
• /
• pp.145-157
• /
• 2003

In this paper, we'd like to briefly introduce Korean common speech database, which project has been started to construct a large scaled speech database since 2002. The project aims at supporting the R&D environment of the speech technology for industries. It encourages domestic speech industries and activates speech technology domestic market. In the first year, the resulting common speech database consists of 25 kinds of databases considering various recording conditions such as telephone, PC, VoIP etc. The speech database will be widely used for speech recognition, speech synthesis, and speaker identification. On the other hand, although the database was originally corrected by manual, still it retains unknown errors and human errors. So, in order to minimize the errors in the database, we tried to find the errors based on the recognition errors and classify several kinds of errors. To be more effective than typical recognition technique, we will develop the automatic error detection method. In the future, we will try to construct new databases reflecting the needs of companies and universities.

• Proceedings of the KSPS conference
• /
• 2002.11a
• /
• pp.81-85
• /
• 2002

We present a statistical analysis of Korean phonological variations using automatic generation of phonetic transcription. We have constructed the automatic generation system of Korean pronunciation variants by applying rules modeling obligatory and optional phonemic changes and allophonic changes. These rules are derived from knowledge-based morphophonological analysis and government standard pronunciation rules. This system is optimized for continuous speech recognition by generating phonetic transcriptions for training and constructing a pronunciation dictionary for recognition. In this paper, we describe Korean phonological variations by analyzing the statistics of phonemic change rule applications for the 60,000 sentences in the Samsung PBS(Phonetic Balanced Sentence) Speech DB. Our results show that the most frequently happening obligatory phonemic variations are in the order of liaison, tensification, aspirationalization, and nasalization of obstruent, and that the most frequently happening optional phonemic variations are in the order of initial consonant h-deletion, insertion of final consonant with the same place of articulation as the next consonants, and deletion of final consonant with the same place of articulation as the next consonants. These statistics can be used for improving the performance of speech recognition systems.

• MALSORI
• /
• no.43
• /
• pp.45-55
• /
• 2002

High reliability of automatic speaker recognition regardless of voice conditions is necessary for forensic application. Audio recordings in real cases are not consistent in voice conditions, such as duration, time interval of recording, given text or conversational speech, transmission channel, etc. In this study the variation of verification error rate of ASR system with the voice conditions was investigated. As a result in order to decrease both false rejection rate and false acception rate, the various voices should be used for training and the duration of train voices should be longer than the test voices.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• v.9 no.2
• /
• pp.469-472
• /
• 2005

This paper propose about effective speech feature parameter that have robust character in effect of noise in realizing speech recognition system. Established MFCC that is the basic parameter used to ASR(Automatic Speech Recognition) and DCTCs that use DCT in basic parameter. Also, proposed delta-Cepstrum and delta-delta-Cepstrum parameter that reconstruct Cepstrum to have information for variation of speech. And compared recognition performance in using HMM. For dimension reduction of each parameter LDA algorithm apply and compared recognition. Results are presented reduced dimension delta-delta-Cepstrum parameter in using LDA recognition performance that improve more than existent parameter in noise environment of various condition.