• ETRI Journal
• /
• 제32권5호
• /
• pp.810-818
• /
• 2010

This paper presents a new framework for integrating untranscribed spoken content into the acoustic training of an automatic speech recognition system. Untranscribed spoken content plays a very important role for under-resourced languages because the production of manually transcribed speech databases still represents a very expensive and time-consuming task. We proposed two new methods as part of the training framework. The first method focuses on combining initial acoustic models using a data-driven metric. The second method proposes an improved acoustic training procedure based on unsupervised transcriptions, in which word endings were modified by broad phonetic classes. The training framework was applied to baseline acoustic models using untranscribed spoken content from parliamentary debates. We include three types of acoustic models in the evaluation: baseline, reference content, and framework content models. The best overall result of 18.02% word error rate was achieved with the third type. This result demonstrates statistically significant improvement over the baseline and reference acoustic models.

• International Journal of Air-Conditioning and Refrigeration
• /
• 제9권1호
• /
• pp.11-19
• /
• 2001

To establish optimum cycle of the inverter-driven heat pump with a variation of frequency, the bypass orifice, which was a short tube haying a bypass hole in the middle, was designed and tested. Flow characteristics of the bypass orifice were measured as a function of orifice geometry and operating conditions. Flow trends with respect to frequency were compared with those of short tube orifices and capillary tubes. Generally, the bypass orifice showed the best flow trends among them. and it would enhance the seasonal energy efficiency ratio of an inverter heat pump system, Based on experimental data, a semi-empirical flow model was developed to predict mass flow rate through bypass orifices. The maximum difference between measured data and model`s prediction was within $\pm$5%.

• 설비공학논문집
• /
• 제11권1호
• /
• pp.109-116
• /
• 1999

To establish optimum cycle of the inverter-driven heat pump with a variation of frequency, the bypass orifice, which is a short tube having a bypass hole in the middle, was designed and tested. Flow characteristics of the bypass orifice were measured as a function of orifice geometry and operating conditions. Flow trends with respect to frequency were compared with those of short tube orifices and capillary tubes. Generally, the bypass orifice showed the best flow trends among them, that will enhance the seasonal energy efficiency ratio of an inverter heat pump system. Based on experimental data, the semi-empirical flow model was developed to predict mass flow rate through bypass orifices. The maximum difference between measured data and model's prediction was within ${\pm}5%$.

• 한국수산과학회지
• /
• 제27권6호
• /
• pp.787-793
• /
• 1994

최적해(optimal solution)를 구하는 최적화기법이 정방해양의 해수순환을 기술하는 간단한 유한차분모델에 대한 부정확한 외력 및 입력 자료 문제에 적용되어짐을 보였다. 수학적 절차는 추정된 자료 및 간단한 순압성 해수순환모델(wind-driven ocean circulation model)을 이용하여 공액경사법에 의한 최적화 방법에 기초하였다. 해는 부정확성 작용에 제약조건을 적용한 목적함수(objective function)를 최적화함으로써 발견되어지며, 이에 대한 부가제약조건(additional constraints) 수의 영향을 살펴보았다.

• 로봇학회논문지
• /
• 제14권4호
• /
• pp.340-347
• /
• 2019

Due to advances in machine intelligence and increased demands for autonomous machines, the complexity of the underlying software platform is increasing at a rapid pace, overwhelming the developers with implementation details. We attempt to ease the burden that falls onto the developers by creating a graphical programming framework we named Splash. Splash is designed to provide an effective programming abstraction for autonomous machines that require stream processing. It also enables programmers to specify genuine, end-to-end timing constraints, which the Splash framework automatically monitors for violation. By utilizing the timing constraints, Splash provides three key language semantics: timing semantics, in-order delivery semantics, and rate-controlled data-driven stream processing semantics. These three semantics together collectively serve as a conceptual tool that can hide low-level details from programmers, allowing developers to focus on the main logic of their applications. In this paper, we introduce the three-language semantics in detail and explain their function in association with Splash's language constructs. Furthermore, we present the internal workings of the Splash programming framework and validate its effectiveness via a lane keeping assist system.

• 대한조선학회논문집
• /
• 제61권4호
• /
• pp.226-235
• /
• 2024

A method for quantifying the adaptability of ship maneuver scenarios for data-driven modeling of ship dynamics is developed based on the principal component analysis. A random maneuver scenario is suggested as a reference for ship dynamics, which can obtain the converged principal components of ship dynamics features by the Monte Carlo simulation. Principal components of conventional maneuver scenarios defined by the International Maritime Organization (IMO) are compared to that of the random maneuver. A conventional ship dynamics model for a container carrier vessel for four degrees of freedom dynamics is introduced to simulate the random and IMO maneuver scenarios. It is confirmed that the IMO tests follow the tendency of random maneuver scenario in terms of execution time and adaptability.

• 한국전산유체공학회지
• /
• 제3권1호
• /
• pp.11-21
• /
• 1998

A vorticity-based method for the numerical solution of the two-dimensional incompressible Navier-Stokes equations is presented. The governing equations for vorticity, velocity and pressure variables are expressed in an integro-differential form. The global coupling between the vorticity and the pressure boundary conditions is fully considered in an iterative procedure when numerical schemes are employed. The finite volume method of the second order TVD scheme is implemented to integrate the vorticity transport equation with the dynamic vorticity boundary condition. The velocity field is obtained by using the Biot-Savart integral. The Green's scalar identity is used to solve the total pressure in an integral approach similar to the surface panel methods which have been well established for potential flow analysis. The present formulation is validated by comparison with data from the literature for the two-dimensional cavity flow driven by shear in a square cavity. We take two types of the cavity now: (ⅰ) driven by non-uniform shear on top lid and body forces for which the exact solution exists, and (ⅱ) driven only by uniform shear (of the classical type).

• Structural Engineering and Mechanics
• /
• 제84권2호
• /
• pp.143-154
• /
• 2022

Corrosion of steel reinforcement is considered as the main cause of concrete structures deterioration, especially those under humid environmental conditions. Hence, fiber reinforced polymer (FRP) bars are being increasingly used as a replacement for conventional steel owing to their non-corrodible characteristics. However, predicting the shear strength of beams reinforced with FRP bars still challenging due to the lack of robust shear theory. Thus, this paper aims to develop an explicit data driven based model to predict the shear strength of FRP reinforced beams using multi-objective evolutionary polynomial regression analysis (MOGA-EPR) as data driven models learn the behavior from the input data without the need to employee a theory that aid the derivation, and thus they have an enhanced accuracy. This study also evaluates the accuracy of predictive models of shear strength of FRP reinforced concrete beams employed by different design codes by calculating and comparing the values of the mean absolute error (MAE), root mean square error (RMSE), mean (𝜇), standard deviation of the mean (𝜎), coefficient of determination (R²), and percentage of prediction within error range of ±20% (a20-index). Experimental database has been developed and employed in the model learning, validation, and accuracy examination. The statistical analysis illustrated the robustness of the developed model with MAE, RMSE, 𝜇, 𝜎, R², and a20-index of 14.6, 20.8, 1.05, 0.27, 0.85, and 0.61, respectively for training data and 10.4, 14.1, 0.98, 0.25, 0.94, and 0.60, respectively for validation data. Furthermore, the developed model achieved much better predictions than the standard predictive models as it scored lower MAE, RMSE, and 𝜎, and higher R² and a20-index. The new model can be used in future with confidence in optimized designs as its accuracy is higher than standard predictive models.

• ETRI Journal
• /
• 제38권5호
• /
• pp.1019-1029
• /
• 2016

The probability hypothesis density (PHD) filter is an effective means to track multiple targets in that it avoids explicit data associations between the measurements and targets. However, the target birth intensity as a prior is assumed to be known before tracking in a traditional target-tracking algorithm; otherwise, the performance of a conventional PHD filter will decline sharply. Aiming at this problem, a novel target birth intensity scheme and an improved measurement-driven scheme are incorporated into the PHD filter. The target birth intensity estimation scheme, composed of both PHD pre-filter technology and a target velocity extent method, is introduced to recursively estimate the target birth intensity by using the latest measurements at each time step. Second, based on the improved measurement-driven scheme, the measurement set at each time step is divided into the survival target measurement set, birth target measurement set, and clutter set, and meanwhile, the survival and birth target measurement sets are used to update the survival and birth targets, respectively. Lastly, a Gaussian mixture implementation of the PHD filter is presented under a linear Gaussian model assumption. The results of numerical experiments demonstrate that the proposed approach can achieve a better performance in tracking systems with an unknown newborn target intensity.

• Biotechnology and Bioprocess Engineering:BBE
• /
• 제11권1호
• /
• pp.67-71
• /
• 2006

This paper compares regression and neural network modeling approaches to predict competitive biosorption equilibrium data. The regression approach is based on the fitting of modified Langmuir-type isotherm models to experimental data. Neural networks, on the other hand, are non-parametric statistical estimators capable of identifying patterns in data and correlations between input and output. Our results show that the neural network approach outperforms traditional regression-based modeling in correlating and predicting the simultaneous uptake of copper and cadmium by a microbial biosorbent. The neural network is capable of accurately predicting unseen data when provided with limited amounts of data for training. Because neural networks are purely data-driven models, they are more suitable for obtaining accurate predictions than for probing the physical nature of the biosorption process.