• ETRI Journal
• /
• v.46 no.1
• /
• pp.48-58
• /
• 2024

This paper presents the development of language tutoring systems for nonnative speakers by leveraging advanced end-to-end automatic speech recognition (ASR) and proficiency evaluation. Given the frequent errors in non-native speech, high-performance spontaneous speech recognition must be applied. Our systems accurately evaluate pronunciation and speaking fluency and provide feedback on errors by relying on precise transcriptions. End-to-end ASR is implemented and enhanced by using diverse non-native speaker speech data for model training. For performance enhancement, we combine semisupervised and transfer learning techniques using labeled and unlabeled speech data. Automatic proficiency evaluation is performed by a model trained to maximize the statistical correlation between the fluency score manually determined by a human expert and a calculated fluency score. We developed an English tutoring system for Korean elementary students called EBS AI Peng-Talk and a Korean tutoring system for foreigners called KSI Korean AI Tutor. Both systems were deployed by South Korean government agencies.

• Journal of the Korea Convergence Society
• /
• v.5 no.3
• /
• pp.17-22
• /
• 2014

The stress between work piece and end mill at the use of end mill at machining and the structural deformation due to the and the pressure are investigated by simulation analysis of three end mill models in this study. These results are achieved with structural and fatigue analyses. Model 1 has the deformation less than model 2 or model 3. As the maximum equivalent stress of model 1 is shown to become the least among all models, model 1 can endure the highest load by comparing with other models. It is useful to estimate the damage prevention and the durability by applying this study result into the design of end mill.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.5
• /
• pp.2499-2522
• /
• 2017

Multi-hop networks are a low-setup-cost solution for enlarging an area of network coverage through multi-hop routing. Carrier sense multiple access with collision avoidance (CSMA/CA) is frequently used in multi-hop networks. Multi-hop networks face multiple problems, such as a rise in contention for the medium, and packet loss under heavy-load, saturated conditions, which consumes more bandwidth due to re-transmissions. The number of re-transmissions carried out in a multi-hop network plays a major role in the achievable quality of service (QoS). This paper presents a statistical, analytical model for the end-to-end delay of contention-based medium access control (MAC) strategies. These strategies schedule a packet before performing the back-off contention for both differentiated heterogeneous data and homogeneous data under saturation conditions. The analytical model is an application of Markov chain theory and queuing theory. The M/M/1 model is used to derive access queue waiting times, and an absorbing Markov chain is used to determine the expected number of re-transmissions in a multi-hop scenario. This is then used to calculate the expected end-to-end delay. The prediction by the proposed model is compared to the simulation results, and shows close correlation for the different test cases with different arrival rates.

• Journal of Communications and Networks
• /
• v.11 no.3
• /
• pp.297-305
• /
• 2009

Multiprotocol label switching (MPLS) networks require dynamic flow admission control to guarantee end-to-end quality of service (QoS) for each Internet protocol (IP) traffic flow. In this paper, we propose to tackle the joint routing and admission control problem for the IP traffic flows in MPLS networks without rerouting already admitted flows. We propose two mathematical programming models for this problem. The first model includes end-to-end delay constraints and the second one, end-to-end packet loss constraints. These end-to-end QoS constraints are imposed not only for the new traffic flow, but also for all already admitted flows in the network. The objective function of both models is to minimize the end-to-end delay for the new flow. Numerical results show that considering end-to-end delay (or packet loss) constraints for all flows has a small impact on the flow blocking rate. Moreover, we reduces significantly the mean end-to-end delay (or the mean packet loss rate) and the proposed approach is able to make its decision within 250 msec.

• Fisheries and Aquatic Sciences
• /
• v.14 no.1
• /
• pp.62-69
• /
• 2011

The effects of actively stimulating devices (ASD) on juvenile flatfish escape were studied to increase escape rates from the cod end by encouraging fish to approach the net wall. Two kinds of ASD were designed: a fluttering net panel, i.e., a free-end flag-like net panel, and a double conical rope array. Escape responses of juvenile bastard halibut were observed in a circulating water channel using two model cod ends, one made with diamond-shaped 43-mm-mesh-size polyethylene (PE) as a high-contrast cod end and the other with polyamid (PA) mono-ply as a low-contrast cod end. Retention rate was significantly lower with the double conical rope-array ASD in the PE cod end than with conventional PE cod ends only or the fluttering net-panel ASD inside the PE cod end. Mean retention rate with the low-contrast PA cod end was also significantly lower than that with the high-contrast PE conventional cod end. Therefore, active fluttering devices using a double conical rope array together with less visible low-contrast netting in the cod end could help to reduce the bycatch of juvenile flatfish by weakening their optomotor response and actively driving fish to the side net panel.

• Transactions of the Korean Society of Machine Tool Engineers
• /
• v.12 no.2
• /
• pp.44-52
• /
• 2003

This research suggests a cutting force model for the ball end milling processes. This model includes the effect of tool run out and tool deflection. In the proposed model, the flutes of ball end mills are considered as series of infinitesimal elements and each cutting edge is assumed to be straight for the analysis of the oblique cutting process, in which the small cutting edge element has been analyzed as an orthogonal cutting process n the plane including the cutting velocity and the chip-flow vector. Therefor, the cutting forces can be calculated through the model using the orthogonal cutting data obtained from the orthogonal cutting test. In order to enhance the performance of the model, the flutes of ball end mill are defined to keep geometric consistency at the peak of the ball part and the junction with the end mill part. The divided infinitesimal cutting edges are regulated to be even lengths. Some experiments show the validity of the developed model in the various cutting coalitions.

• Journal of the Korean Society for Precision Engineering
• /
• v.14 no.4
• /
• pp.38-45
• /
• 1997

A new dynamic model for the cutting process inb the end milling process is developed. This model, which describes the dynamic response of the end mill, the chip load geometry including tool runout, the dependence of the cutting forces on the chip load, is used to predict the dynamic cutting force during the end milling process. In order to predict accurately cutting forces and tool vibration, the model which uses instantaneous specific cutting force, inclueds both regenerative effect and penetration effect, The model is verified through comparisons of model predicted cutting force with measured cutting force obtained from machining experiments.

• Steel and Composite Structures
• /
• v.15 no.1
• /
• pp.81-101
• /
• 2013

In this paper a robust 2-noded connection element has been developed for modelling the bolted end-plate connection between steel beam and column at elevated temperatures. The numerical procedure described is based on the model proposed by Huang (2011), incorporating additional developments to more precisely determinate the tension, compression and bending moment capacities of end-plate connection in fire. The proper failure criteria are proposed to calculate the tension capacity for each individual bolt row. In this new model the connection failure due to bending, axial tension, compression and shear are considered. The influence of the axial force of the connected beam on the connection is also taken into account. This new model has the advantages of both the simple and component-based models. In order to validate the model a total of 22 tests are used. It is evident that this new connection model has ability to accurately predict the behaviour of the end-plate connection at elevated temperatures, and can be used to represent the end-plate connections in supporting performance-based fire resistance design of steel-framed composite buildings.

• Proceedings of the Korea Concrete Institute Conference
• /
• 2000.10a
• /
• pp.235-240
• /
• 2000

Although dapped-end beams are widely used in bridge and building structures, there are not any specific and reasonable design regulations on dapped-end beams. In this study, the validity of the suggested experimental and empirical design methods, conventional strut-tie model approach, and nonlinear strut-tie model approach is evaluated through the analysis of dapped-end beams tested to failure. The nonlinear strut-tie model approach proved to be the most suitable method for dapper-end beam design.

• Fisheries and Aquatic Sciences
• /
• v.11 no.1
• /
• pp.61-69
• /
• 2008

Using empirical data of fish performance and physiological limits as well as physical stimuli and environmental data, a cod-end selectivity model based on a chaotic behavior model using the psycho-hydraulic wheel and neural-network approach was established to predict fish escape or herding responses in trawl and cod-end designs. Fish responses in the cod-end were categorized as escape or herding reactions based on their relative positions and reactions to the net wall. Fish movements were regulated by three factors: escape time, a visual looming effect, and an index of body girth-mesh size. The model was applied to haddock in a North Sea bottom trawl including frequencies of movement components, swimming speed, angular velocity, distance to net wall, and the caught-fish ratio; simulation results were similar to field observations. The ratio of retained fish in the cod-end was limited to 37-95% by optomotor coefficient values of 0.3-1.0 and to 13-67% by looming coefficient values of 0.1-1.0. The selectivity curves generated by this model were sensitive to changes in mesh size, towing speed, mesh type, and mesh shape.