• The Journal of the Acoustical Society of Korea
• /
• v.40 no.5
• /
• pp.530-536
• /
• 2021

Recently, due to the development of deep learning, end-to-end speech recognition, which directly maps graphemes to speech signals, shows good performance. Especially, among the end-to-end models, conformer shows the best performance. However end-to-end models only focuses on the probability of which grapheme will appear at the time. The decoding process uses a greedy search or beam search. This decoding method is easily affected by the final probability output by the model. In addition, the end-to-end models cannot use external pronunciation and language information due to structual problem. Therefore, in this paper conformer with lexicon transducer is proposed. We compare phoneme-based model with lexicon transducer and grapheme-based model with beam search. Test set is consist of words that do not appear in training data. The grapheme-based conformer with beam search shows 3.8 % of CER. The phoneme-based conformer with lexicon transducer shows 3.4 % of CER.

• Journal of the Korean Chemical Society
• /
• v.54 no.3
• /
• pp.275-282
• /
• 2010

2-Fluorocyclopropanemethanol and 2-chlorocyclopropanemethanol have been studied with MP2 and B3LYP methods with 6-311++G(d,p) basis set. The optimized structures show several stable conformers. The most stable conformer show the possibility of intramolecular hydrogen bonding, but the distance between $H{\cdots}F$, or $H{\cdots}Cl$ is longer than van der Waals radii and it may not be strong covalent bonding. Rather the second stable conformer has optimum structure for intramolecular hydrogen bonding but the energy of the conformer is 5 ~ 7 kJ higher than the most stable conformer. When the methanol group and the F or Cl atom have opposite direction, the conformers are less stable than the most stable conformer.

• The Journal of the Acoustical Society of Korea
• /
• v.42 no.3
• /
• pp.214-220
• /
• 2023

In this paper, we propose a method to improve texture recognition performance from haptic acceleration signals representing the texture characteristics of object surface materials by using a Conformer model that combines the advantages of a convolutional neural network and a transformer. In the proposed method, three-axis acceleration signals generated by impact sound and vibration are combined into one-dimensional acceleration data while a person contacts the surface of the object materials using a tool such as a stylus , and the logarithmic Mel-spectrogram is extracted from the haptic acceleration signal similar to the audio signal. Then, Conformer is applied to the extracted the logarithmic Mel-spectrogram to learn main local and global frequency features in recognizing the texture of various object materials. Experiments on the Lehrstuhl für Medientechnik (LMT) haptic texture dataset consisting of 60 materials to evaluate the performance of the proposed model showed that the proposed method can effectively recognize the texture of the object surface material better than the existing methods.

• The Journal of the Acoustical Society of Korea
• /
• v.41 no.3
• /
• pp.359-366
• /
• 2022

In this paper, we present the Korean menu-ordering Sentence Text-to-Speech (TTS) system using conformer-based FastSpeech2. Conformer is the convolution-augmented transformer, which was originally proposed in Speech Recognition. Combining two different structures, the Conformer extracts better local and global features. It comprises two half Feed Forward module at the front and the end, sandwiching the Multi-Head Self-Attention module and Convolution module. We introduce the Conformer in Korean TTS, as we know it works well in Korean Speech Recognition. For comparison between transformer-based TTS model and Conformer-based one, we train FastSpeech2 and Conformer-based FastSpeech2. We collected a phoneme-balanced data set and used this for training our models. This corpus comprises not only general conversation, but also menu-ordering conversation consisting mainly of loanwords. This data set is the solution to the current Korean TTS model's degradation in loanwords. As a result of generating a synthesized sound using ParallelWave Gan, the Conformer-based FastSpeech2 achieved superior performance of MOS 4.04. We confirm that the model performance improved when the same structure was changed from transformer to Conformer in the Korean TTS.

• Bulletin of the Korean Chemical Society
• /
• v.31 no.2
• /
• pp.457-460
• /
• 2010

The geometrical structures and energetics of positively doubly charged fullerene dimer $(C_{60})_2{^{2+}}$ conformers were studied using semiempirical PM3 and MNDO, Hartree-Fock (HF), and Hybrid B3LYP density functional methods. The shape of the HOMO-LUMO for the three conformers was also analyzed. The gauche conformer was the most stable of the three conformers. The anti conformer was more stable than the syn conformer.

• Bulletin of the Korean Chemical Society
• /
• v.29 no.7
• /
• pp.1374-1378
• /
• 2008

Using DFT B3LYP/6-31+G(d,p)//B3LYP/6-31G(d,p) calculation method, stable molecular structures were optimized for the p-tert-butylcalix[4]arene functionalized at lower rim by cage-annulated crown ether (1) in two different conformers and their potassium-ion complexes. Cone conformer of free host 1 was slightly more stable than partial-cone conformer. For two different kinds of complexation mode, the potassium ion in benzene-rings (bz) pocket showed comparable complexation efficiency with the cation in cage-annulated crown-ether (cr) for the cone and partial-cone conformers of 1. The complex (1${\bullet}K^+$) in the cr-binding mode for the partial-cone conformer was more stable than the cone conformer for B3LYP/6-31G(d,p) geometry optimization. However, $1_{(cone)}{\bullet}K^+$(cr) showed lower single-point energy than the $1_{(pc)}{\bullet}K^+$(cr) for B3LYP/6- 31+G(d,p) calculation method.

• The Journal of the Acoustical Society of Korea
• /
• v.40 no.5
• /
• pp.488-495
• /
• 2021

We propose a speech recognition system based on conformer. Conformer is known to be convolution-augmented transformer, which combines transfer model for capturing global information with Convolution Neural Network (CNN) for exploiting local feature effectively. The baseline system is developed to be a transfer-based speech recognition using Long Short-Term Memory (LSTM)-based language model. The proposed system is a system which uses conformer instead of transformer with transformer-based language model. When Electronics and Telecommunications Research Institute (ETRI) speech corpus in AI-Hub is used for our evaluation, the proposed system yields 5.7 % of Character Error Rate (CER) while the baseline system results in 11.8 % of CER. Even though speech corpus is extended into other domain of AI-hub such as NHNdiguest speech corpus, the proposed system makes a robust performance for two domains. Throughout those experiments, we can prove a validation of the proposed system.

• Bulletin of the Korean Chemical Society
• /
• v.25 no.4
• /
• pp.452-459
• /
• 2004

The geometrical structures of various isomers of hexafluoro-1,3-butadiene (HFBD), tetrafluoro-1,3-butadiene (TFBD), and difluoro-1,3-butadiene (DFBD) have been studied theoretically. Natural steric and natural resonance theory (NRT) analyses indicate that the lower energy of skew s-cis conformer of hexafluoro-1,3-butadiene than that of the s-trans conformer is originated from the strong steric repulsions between fluorine atoms particularly in the s-trans conformer. The resonance structures generated by NRT also show that the lone electron pairs of fluorine atoms effectively extend the conjugation, and the large differences in energy among the structural isomers of tetrafluoro-1,3-butadiene and difluoro-1,3-butadiene are in part attributed to the differences in the delocalization energies, in addition to the steric repulsion between fluorine atoms. Other interatomic interactions, such as hydrogen bonding, also play important roles in determination of the structures of isomers of tetrafluoro-1,3-butadiene and difluoro-1,3-butadiene.

• Bulletin of the Korean Chemical Society
• /
• v.32 no.1
• /
• pp.208-212
• /
• 2011

In this work, the calculated nuclear quadrupole coupling constants of $^{17}O$ in some styrylquinoline conformers were presented. The calculations were carried out to find the relationships between the charge distribution of styrylquinolines and their pharmaceutical behavior and to explore the differences among the electronic structures of some conformers of these potent HIV IN inhibitors. Furthermore, the HIV IN inhibitory of R1 and R2 rotamers was compared. On the basis of our results: - Charge density on oxygen atoms of carboxyl moiety has a dominant role in the drug activity. - The a conformer in which a divalent hydrogen atom is a link, has more capability in antiviral drug treatment. - The R1 conformer, as a $Mg^{+2}$ chelating agent, is better than R2 conformer and thus it is more inhibitor of HIV IN.

• Bulletin of the Korean Chemical Society
• /
• v.8 no.1
• /
• pp.39-45
• /
• 1987

Daunomycin, an anthracycline antibiotic, has been found to inhibit virus multiplication and shows considerable activity against tumors. Its activity may be varied by conformational changes of daunomycin. The conformational changes are come from the pucker of D-ring and variation of environments. We have carried out conformational analyses by using empirical potential function. We found that when daunomycin is hydrated or bound to $Mg^{2+}$ ion, the minimum conformer of each state is altered from ${\alpha}$ conformer to ${\beta}$ conformer through the pathway having four local minima. Our calculated results are in good agreements with those of X-ray crystallography and biological experiments, in which metal ion inhibits the binding of daunomycin to DNA.