• MALSORI
• /
• no.51
• /
• pp.151-165
• /
• 2004

For educational and research purposes, a Korean speech recognition platform is designed. It is based on an object-oriented architecture and can be easily modified so that researchers can readily evaluate the performance of a recognition algorithm of interest. This platform will save development time for many who are interested in speech recognition. The platform includes the following modules: Noise reduction, end-point detection, met-frequency cepstral coefficient (MFCC) and perceptually linear prediction (PLP)-based feature extraction, hidden Markov model (HMM)-based acoustic modeling, n-gram language modeling, n-best search, and Korean language processing. The decoder of the platform can handle both lexical search trees for large vocabulary speech recognition and finite-state networks for small-to-medium vocabulary speech recognition. It performs word-dependent n-best search algorithm with a bigram language model in the first forward search stage and then extracts a word lattice and restores each lattice path with a trigram language model in the second stage.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1998.06c
• /
• pp.47-50
• /
• 1998

본 논문은 자동차의 편의성 및 안전성의 동시 확보를 위하여, 보조적 스위치의 조작없이 상시 음성의 입,출력이 가능하도록 하며, band pass filter를 이용하여 잡음환경에서 자동으로 정확하게 음성구간 검출(End Point Detection)을 하게 하였다. Reference Pattern은 Dynamic Multi-Section(DMS)[1] 모델을 사용하였고 차량의 속도에 따라 자동으로 잡음환경에 강인한 모델을 선택하도록 하였으며, 음성의 특징 파라미터와 인식 알고리즘은 Perceptual Linear Predictive(PLP) 13차와 One Stage Dynamic Programming(OSDP)를 사용하였다. 주행중인 자동차 환경(30~70km/h)에서 자주 사용되는 차량제어 명령 33개에 대하여 화자독립 92.98%, 화자종속 94.44% 인식율을 구하였다. 또한 주행중인 차량에서 카폰, 핸드폰 사용으로 인한 사고를 줄이기 위하여 음성으로 전화를 걸 수 있도록 하는 Voice Dialing 기능도 구현하였다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.14 no.2
• /
• pp.201-206
• /
• 2004

This study attempted to develop a speech recognition module applied to a wheelchair for the physically handicapped. In the proposed speech recognition module, TMS320C32 was used as a main processor and Mel-Cepstrum 12 Order was applied to the pro-processor step to increase the recognition rate in a noisy environment. DTW (Dynamic Time Warping) was used and proven to be excellent output for the speaker-dependent recognition part. In order to utilize this algorithm more effectively, the reference data was compressed to 1/12 using vector quantization so as to decrease memory. In this paper, the necessary diverse technology (End-point detection, DMA processing, etc.) was managed so as to utilize the speech recognition system in real time

• Applied Science and Convergence Technology
• /
• v.23 no.6
• /
• pp.328-339
• /
• 2014

In this article, plasma monitoring tools and mulivariate analysis techniques were reviewed. Optical emission spectroscopy was reviewed for a chemical composition analysis tool and RF V-I probe for a physical analysis tool for plasma monitoring. Multivariate analysis techniques are discussed to the sensitivity improvement. Principal component analysis (PCA) is one of the widely adopted multivariate analysis techniques and its application to end-point detection of plasma etching process is discussed.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2001.11b
• /
• pp.189-192
• /
• 2001

Recently, STI process is getting attention as a necessary technology for making high density of semiconductor by devices isolation method. However, it does have various problems caused by CMP nprocess, such as torn oxide defects, nitride residues on oxide, damages of si active region, contaminations due to post-CMP cleaning, difficulty of accurate end point detection in CMP process, etc. In this work, the various defects induced by CMP process was introduced and the above mentioned problems of CMP process was examined in detail. Finally, the guideline of future CMP process was presented to reduce the effects of these defects.

• The Transactions of the Korean Institute of Electrical Engineers C
• /
• v.52 no.10
• /
• pp.441-446
• /
• 2003

The chemical mechanical polishing (CMP) process is now widely employed in the ultra large scale integrated (ULSI) semiconductor fabrication. Especially, shallow trench isolation (STI) has become a key isolation scheme for sub-0.13/0.10${\mu}{\textrm}{m}$ CMOS technology. The most important issues of STI-CMP is to decrease the various defects such as nitride residue, dishing, and tom oxide. To solve these problems, in this paper, we studied the planarization characteristics using slurry additive with the high selectivity between $SiO_2$ and $Si_3$$N_4$ films for the purpose of process simplification and in-situ end point detection. As our experimental results, it was possible to achieve a global planarization and STI-CMP process could be dramatically simplified. Also, we estimated the reliability through the repeated tests with the optimized process conditions in order to identify the reproducibility of STI-CMP process.

• Proceedings of the KWS Conference
• /
• 2005.06a
• /
• pp.326-328
• /
• 2005

High speed rotational arc sensing is an important method to detect the torch deviation during automatic seam tracking of arc welding. In this paper, a mathematic model of high speed rotational arc sensing is analyzed. The simulation results are consistent with the experimental results. The current waveforms at the beginning of the welding are different from those at middle of the welding because of the formation of the weld bead profile. The signal patterns for seam tracking and end-point detection are proposed. A phase shift between the rotation and the current variation is also discovered in the experiments.

• Proceedings of the KOSOMBE Conference
• /
• v.1993 no.05
• /
• pp.90-93
• /
• 1993

In this study, Korean speech was recognized by using spatio-temporal recognition neural network. The subjects of speech are numeric speech from zero to nine and basic command which might be used for motorized wheelchair developed it own Lab. Rabiner and Sambur's method of speech detection was used in determining end-point of speech, speech parameter was extracted by using LPC 16 order. The recognition rate was over 90%.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 1999.05a
• /
• pp.211-213
• /
• 1999

In this paper, we documented the controlling oxide removal amount on the pattern wafer using removal rate and removal thickness of blanket wafer. There was the strong correlation relationship for both(correlation factor:0.7109). So, we could confirm the repeatability as applying for STI CMP process from the obtained linear formular. As the result of repeatability test, the difference of calculated polishing time and actual polishing time was 3.48 seconds based on total 50 lots. If this time is converted into the thickness, it is from 104$\AA$ to 167$\AA$. It is possible to be ignored because it is under the process margin.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 1999.05a
• /
• pp.501-504
• /
• 1999

In the shallow trench isolation(STI) chemical mechanical polishing(CMP) process, the key issues are the optimized thickness control within- wafer-non-uniformity, and the possible defects such as nitride residue and pad oxide damage. These defects after STI CMP process were discussed to accomplish its optimum process condition. To understand its optimum process condition, overall STI related processes including reverse moat etch, trench etch, STI filling and STI CMP were discussed. It is represented that the nitride residue can be occurred in the condition of high post CMP thickness and low trench depth. In addition there are remaining oxide on the moat surface after reverse moat etch. It means that reverse moat etching process can be the main source of nitride residue. Pad oxide damage can be caused by over-polishing and high trench depth.