• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.559-560
• /
• 2018

순환 신경망(RNN) 기반의 Long Short-Term Memory(LSTM)는 자연어처리 분야에서 우수한 성능을 보이는 모델이다. 음성을 문자로 변환해주는 Speech to Text (STT)를 이용해 자막을 생성하고, 생성된 자막을 다른 언어로 동시에 번역을 해주는 서비스가 활발히 진행되고 있다. STT를 사용하여 자막을 추출하는 경우에는 마침표가 없이 전부 연결된 문장이 생성되기 때문에 정확한 번역이 불가능하다. 본 논문에서는 영어자막의 자동 번역 시, 정확도를 높이기 위해 텍스트를 문장으로 분할하여 마침표를 생성해주는 방법을 제안한다. 이 때, LSTM을 이용하여 데이터를 학습시킨 후 테스트한 결과 62.3%의 정확도로 마침표의 위치를 예측했다.

• 제어로봇시스템학회:학술대회논문집
• /
• 1996.10b
• /
• pp.1292-1295
• /
• 1996

This paper proposes an adaptive control technique for the autopilot design of STT missile. Dynamics of the missile is highly nonlinear and the equilibrium point is vulnerable to change due to fast maneuvering. Therefore nonlinear control techniques are desirable for the autopilot design of the missile. The nonlinear controller requires the exact model to obtain satisfactory performance. Generally a look-up table is used for the dynamic coefficients of a missile, so there must be coefficients error during actual flight, and the performance of the nonlinear controller using these data can be degraded. The proposed adaptive control technique compensates the nonlinear controller with modeling error resulting from the error of aerodynamic data and disturbance. To investigate the usefulness, the proposed method is applied to autopilot design of STT missile through simulations.

• The Korean Journal of Physiology and Pharmacology
• /
• v.17 no.6
• /
• pp.553-558
• /
• 2013

Spinal dorsal horn nociceptive neurons have been shown to undergo long-term synaptic plasticity, including long-term potentiation (LTP) and long-term depression (LTD). Here, we focused on the spinothalamic tract (STT) neurons that are the main nociceptive neurons projecting from the spinal cord to the thalamus. Optical technique using fluorescent dye has made it possible to identify the STT neurons in the spinal cord. Evoked fast mono-synaptic, excitatory postsynaptic currents (eEPSCs) were measured in the STT neurons. Time-based tetanic stimulation (TBS) was employed to induce long-term potentiation (LTP) in the STT neurons. Coincident stimulation of both pre- and postsynaptic neurons using TBS showed immediate and persistent increase in AMPA receptor-mediated EPSCs. LTP can also be induced by postsynaptic spiking together with pharmacological stimulation using chemical NMDA. TBS-induced LTP observed in STT neurons was blocked by internal BAPTA, or $Ni^{2+}$, a T-type VOCC blocker. However, LTP was intact in the presence of L-type VOCC blocker. These results suggest that long-term plastic change of STT neurons requires NMDA receptor activation and postsynaptic calcium but is differentially sensitive to T-type VOCCs.

• The Journal of Korean Physical Therapy
• /
• v.28 no.2
• /
• pp.142-148
• /
• 2016

Purpose: The aim of this pilot study was to determine the effect of soft tissue technique (STT) in Axillary Web Syndrome (AWS) following sentinel Lymph Node Biopsy in breast cancer patients by examining the upper extremity function, range of motion, and pain. Methods: Nineteen patients with breast cancer-related AWS were evaluated. STT was performed on the symptom area for treatment of AWS symptoms. We evaluated AWS symptoms and pain intensity using a visual analogue scale (VAS), and functional disability using the Korean version of the Disabilities of Arm, Shoulder, and Hand (DASH) questionnaire, glenohumeral joint (GHJ) flexion and abduction range of motion (ROM) with or without elbow extension. Results: Visual analyses of the data suggest a modest effect of STT in improving GHJ flexion, abduction ROM with or without elbow extension, DASH for upper extremity function, and Pain. The statistically significant improvement in baseline observed for pain, DASH, and ROM data made it impossible to assess the effects of STT on those outcomes. There were no adverse events. Conclusions: STT may be an effective and safe treatment option for AWS patients recovering from breast cancer treatment; however, further research is needed.

• Journal of the Korean Magnetics Society
• /
• v.19 no.1
• /
• pp.22-27
• /
• 2009

Reduction of the critical current density ($J_c$) for STT magnetization switching is most important issue of magnetic tunnel junctions (MTJs) based MRAM. This report describes how to decrease the Jc and will introduce the recent research progresses of STT-MRAM devices with material engineering and structural improvement, respectively.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.716-721
• /
• 2023

Inverse Text Normalization(ITN) 모델은 음성 인식(STT) 엔진의 중요한 후처리 영역 중 하나이며, STT 인식 결과의 가독성을 개선한다. 최근 ITN 모델에 심층신경망을 활용한 연구가 진행되고 있다. 심층 신경망을 사용하는 대부분의 선행연구는 문장 내 변환이 필요한 부분에 토큰 태깅을 진행하는 방식이다. 그러나 이는 Out-of-vocabulary(OOV) 이슈가 있으며, 학습 데이터 구축 시 토큰 단위의 섬세한 태깅 작업이 필요하다는 한계점이 존재한다. 더불어 선행 연구에서는 STT 인식 결과를 그대로 사용하는데, 이는 띄어쓰기가 중요한 한국어 ITN 처리에 변환 성능을 보장할 수 없다. 본 연구에서는 BART 기반 생성 모델로 생성형 ITN 모델을 구축하였고, Bi-LSTM 기반 보조 신경망 모델을 결합하여 STT 인식 결과에 대한 고유명사 처리, 띄어쓰기 교정 기능을 보완한 모델을 제안한다. 또한 보조 신경망을 통해 생성 모델 처리 여부를 판단하여 평균 추론 속도를 개선하였다. 실험을 통해 두 모델의 각 정량 성능 지표에서 우수한 성능을 확인하였고 결과적으로 본 연구에서 제안하는 두 모델의 결합된 방법론의 효과성을 제시하였다.

• Journal of Magnetics
• /
• v.17 no.2
• /
• pp.73-77
• /
• 2012

We investigate spin transfer torque (STT) in magnetic multilayer structures using micromagnetic simulations. We implement the STT contribution for magnetic multilayer structures in addition to the Landau-Lifshitz-Gilbert (LLG) micromagnetic simulators. In addition to the Sloncewski STT term, the zero, first, and second order field-like terms are also considered as well as the effects of the Oersted field due to the running current are addressed. We determine the switching current densities of the free layer with the exchange biased synthetic ferrimagnetic reference layers for various cases.

• 제어로봇시스템학회:학술대회논문집
• /
• 1992.10a
• /
• pp.412-417
• /
• 1992

In this paper, we present a method to design a coupled autopilot for STT missiles which have severe aerodynamic cross-coupling. The aerodynamic model is derived in the meneuver plane and, based on that model, an autopilot scheduled by the normal acceleration and the estimated bank angle is designed. Bank angle is obtained by a simple estimator. With the proposed autopilot, it is shown by computer simulations that induced moments are properly compensated and the performance is supiorior to the conventional autopilot.

• JSTS:Journal of Semiconductor Technology and Science
• /
• v.14 no.6
• /
• pp.728-732
• /
• 2014

Macro-model of magnetic tunnel junction (MTJ) for spin transfer torque magnetic random access memory (STT-MRAM) has been developed. The macro-model can describe the dynamic behavior such as the state change of MTJ as a function of the pulse width of driving current and voltage. The statistical behavior has been included in the model to represent the variation of the MTJ characteristic due to process variation. The macro-model has been developed in Verilog-A.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.13 no.3
• /
• pp.125-131
• /
• 2018

STT-RAM is attracting as a next generation Non-volatile memory for replacing cache memory with low leakage energy, high integration and memory access performance similar to SRAM. However, there is problem of write operations as the other Non_volatile memory. Hybrid cache memory using SRAM and STT-RAM is attracting attention as a cache memory structure with lowe power consumption. Despite this, reducing the leakage energy consumption by the STT-RAM is still lacking access to the Dynamic energy. In this paper, we proposed as energy management method such as a way-selection approach for hybrid L2 cache fo SRAM and STT-RAM and memory selection method of write/read operation. According to the simulation results, the proposed hybrid cache memory reduced the average energy consumption by 40% on SPEC CPU 2006, compared with SRAM cache memory.