Advanced Information Networking and Applications, 2019
Data mining information using various indices and determining candidates that can be judged as ha... more Data mining information using various indices and determining candidates that can be judged as having the same tendency based on similarity between documents is common. The accuracy of similarity largely depends on a sufficient amount of data and requires advanced analysis using natural language processing. In this paper, we present an approach for filtering based on the cosine similarity graph and clustering between candidates. For filtering candidates, we focus on the inflection point of the graphs when plotting by sorting similarities in descending order. By clustering among higher similarities, we aim at filtering candidates that cannot be removed without analyzing advanced natural language processing. The proposed method was applied to movie reviews and sightseeing location reviews written in Japanese. Although this study is a work in progress, it shows that candidates can be recommended without having to manually apply natural language processing, such as preparing stopwords for each category.
Advanced Information Networking and Applications, 2019
Using a blockchain, an application that guarantees reliability, can be realized by managing a dis... more Using a blockchain, an application that guarantees reliability, can be realized by managing a distributed ledger of agreements. A blockchain’s usefulness in applications requiring trust, such as virtual currency, has been confirmed, and it is being used in actual services. However, the time taken to add data to the distributed ledger and the cost of this operation may be too high. As a result, other real applications for blockchains have not been found yet. In this research, we design and implement an application for lending and borrowing books on a distributed ledger. The application is implemented as a distributed application on Ethereum, which is a platform used for building blockchain-based smart contracts. Users of this application can lend and borrow books through a web application that functions as the user interface of the application. The usage logs show that the time and fees required for adding data to the distributed ledger are acceptable.
2013 IEEE Global Engineering Education Conference (EDUCON), 2013
ABSTRACT In this paper, we describe a case study exploring two topics: understanding students&... more ABSTRACT In this paper, we describe a case study exploring two topics: understanding students' effective behaviors for e-learning and investigating the different types of intrinsic and extrinsic motivation behind student behaviors. A series of individual data were acquired from a learning management system, e-learning system, and qualification exam results, and statistically analyzed. The results demonstrated that students who completed the e-learning contents scored higher in the qualification exam and had a stronger intrinsic motivation to acquire self-respect (gaining confidence by passing the exam), even if they had less interest in programming.
Proceedings of the 18th International Conference on Intelligent Virtual Agents, 2018
We present a novel framework to automatically generate natural gesture motions accompanying speec... more We present a novel framework to automatically generate natural gesture motions accompanying speech from audio utterances. Based on a Bi-Directional LSTM Network, our deep network learns speech-gesture relationships with both backward and forward consistencies over a long period of time. Our network regresses a full 3D skeletal pose of a human from perceptual features extracted from the input audio in each time step. Then, we apply combined temporal filters to smooth out the generated pose sequences. We utilize a speech-gesture dataset recorded with a headset and marker-based motion capture to train our network. We validated our approach with a subjective evaluation and compared it against "original" human gestures and "mismatched" human gestures taken from a different utterance. The evaluation result shows that our generated gestures are significantly better than the "mismatched" gestures with respect to time consistency. The generated gesture also shows marginally significant improvement in terms of semantic consistency when compared to "mismatched" gestures.
Communications in Computer and Information Science, 2017
In our research, we recorded 298 min (1049 sentences) of speech audio data and the motion capture... more In our research, we recorded 298 min (1049 sentences) of speech audio data and the motion capture data of the accompanying gestures from two 25-year-old male participants aiming for future usage in deep learning concerning gesture and speech. The data was recorded in form of an interview, the participant explaining a topic prepared in advance, using a headset microphone and the motion capture softwarë Motivë. The speech audio was stored in mp3, and the motion data was stored in bvh, related as data from the same sentence. We aimed to mainly acquire metaphoric gestures and iconic, as categorized by McNiel. For the categories of the recorded gestures, metaphoric gestures appeared the most, 68.41% of all gestures, followed by 23.73% beat gestures, 4.76% iconic gestures, and 3.11% deictic gestures.
The Proceedings of Design & Systems Conference, 2004
Eng 血 eeh 皿 g supPor 仁 system fl) r medlanical design based on the tedhnology of Web Servioes 捻 p... more Eng 血 eeh 皿 g supPor 仁 system fl) r medlanical design based on the tedhnology of Web Servioes 捻 propesed and implemented as an 血 tegrated design system by CADICAMICAE 血 disthbuted envirx)mnellt ． Knowledge base system for mechanical design and numerical analysis can heip fo organize digital processes f (｝ r design by analys 追 ． Un 丗 ed 血 」 brmation of 丘 gure or ana 】 ysB and p買)oess 血g data by XML expwession can be used 血 each design process under physi 〔 nl testriCtions ． Web SelVices can proVide high end computing system for the eng 血 eeri 皿 g enVil ' enment for皿 ore mechanical engtiieer aiid show the possibdity ofintegratedmodel and eXperimental system as a basic system ofproduction knowledge．
Electronic and Photonic Packaging, Electrical Systems Design and Photonics, and Nanotechnology, 2002
In development of surface mount technology for Ball Grid Array (BGA) or flip-chip assemblies, it ... more In development of surface mount technology for Ball Grid Array (BGA) or flip-chip assemblies, it is important to reduce stress concentration in solder joints as it is immediately effective in improving the fatigue life of the assembly. Thus, the ability to predict and control the joint geometries is critical to obtaining robust and reliable designs of interconnects. Other than the issue of the bump shape, there are many problems concerning soldering such as bridging or self-alignment in which surface tension is definitely involved. This paper attempts to apply the FEM approach in solving these problems. Rigid-plastic FEM which is based on iteration for the velocity field in an incompressible viscous fluid is an approach to large deformation analysis. The flow stress is described by the viscosity of the fluid and the strain rate. By introducing an automatic mesh updating procedure, transient problems with free boundary can be treated. We applied this concept to the prediction of sold...
2015 IEEE Global Engineering Education Conference (EDUCON), 2015
Speech-based test is a one of the best ways to train students' level of proficiency. However,... more Speech-based test is a one of the best ways to train students' level of proficiency. However, in e-learning it is difficult for students to maintain their concentration when speaking to a monitor. We propose speech-based online test system employing a human-likely embodied graphical agent as a listener. The developed system consists of three modules, user recognition, agent's non-verbal behavior realization, and learning controller. The system is developed with HTML5, Javascript, and WebGL, so that it should be compatible with any modern Web-browsers. To investigate the effects of the use of the listener agent in speech-based test on learning efficacy and naturalness of speech, we will conduct a one-way between-subject design experiment with three groups. In experimental group, participants speech to a listening agent with listener behavior depicted on a monitor during a speech-based test. On the other hand, in first control group, participants take the test with the listening agent but the agent does not move while the participants' speech. In second control group, participants take the test without the listening agent. Then, the result of listening scores will be statistically tested.
Advanced Information Networking and Applications, 2019
Data mining information using various indices and determining candidates that can be judged as ha... more Data mining information using various indices and determining candidates that can be judged as having the same tendency based on similarity between documents is common. The accuracy of similarity largely depends on a sufficient amount of data and requires advanced analysis using natural language processing. In this paper, we present an approach for filtering based on the cosine similarity graph and clustering between candidates. For filtering candidates, we focus on the inflection point of the graphs when plotting by sorting similarities in descending order. By clustering among higher similarities, we aim at filtering candidates that cannot be removed without analyzing advanced natural language processing. The proposed method was applied to movie reviews and sightseeing location reviews written in Japanese. Although this study is a work in progress, it shows that candidates can be recommended without having to manually apply natural language processing, such as preparing stopwords for each category.
Advanced Information Networking and Applications, 2019
Using a blockchain, an application that guarantees reliability, can be realized by managing a dis... more Using a blockchain, an application that guarantees reliability, can be realized by managing a distributed ledger of agreements. A blockchain’s usefulness in applications requiring trust, such as virtual currency, has been confirmed, and it is being used in actual services. However, the time taken to add data to the distributed ledger and the cost of this operation may be too high. As a result, other real applications for blockchains have not been found yet. In this research, we design and implement an application for lending and borrowing books on a distributed ledger. The application is implemented as a distributed application on Ethereum, which is a platform used for building blockchain-based smart contracts. Users of this application can lend and borrow books through a web application that functions as the user interface of the application. The usage logs show that the time and fees required for adding data to the distributed ledger are acceptable.
2013 IEEE Global Engineering Education Conference (EDUCON), 2013
ABSTRACT In this paper, we describe a case study exploring two topics: understanding students&... more ABSTRACT In this paper, we describe a case study exploring two topics: understanding students' effective behaviors for e-learning and investigating the different types of intrinsic and extrinsic motivation behind student behaviors. A series of individual data were acquired from a learning management system, e-learning system, and qualification exam results, and statistically analyzed. The results demonstrated that students who completed the e-learning contents scored higher in the qualification exam and had a stronger intrinsic motivation to acquire self-respect (gaining confidence by passing the exam), even if they had less interest in programming.
Proceedings of the 18th International Conference on Intelligent Virtual Agents, 2018
We present a novel framework to automatically generate natural gesture motions accompanying speec... more We present a novel framework to automatically generate natural gesture motions accompanying speech from audio utterances. Based on a Bi-Directional LSTM Network, our deep network learns speech-gesture relationships with both backward and forward consistencies over a long period of time. Our network regresses a full 3D skeletal pose of a human from perceptual features extracted from the input audio in each time step. Then, we apply combined temporal filters to smooth out the generated pose sequences. We utilize a speech-gesture dataset recorded with a headset and marker-based motion capture to train our network. We validated our approach with a subjective evaluation and compared it against "original" human gestures and "mismatched" human gestures taken from a different utterance. The evaluation result shows that our generated gestures are significantly better than the "mismatched" gestures with respect to time consistency. The generated gesture also shows marginally significant improvement in terms of semantic consistency when compared to "mismatched" gestures.
Communications in Computer and Information Science, 2017
In our research, we recorded 298 min (1049 sentences) of speech audio data and the motion capture... more In our research, we recorded 298 min (1049 sentences) of speech audio data and the motion capture data of the accompanying gestures from two 25-year-old male participants aiming for future usage in deep learning concerning gesture and speech. The data was recorded in form of an interview, the participant explaining a topic prepared in advance, using a headset microphone and the motion capture softwarë Motivë. The speech audio was stored in mp3, and the motion data was stored in bvh, related as data from the same sentence. We aimed to mainly acquire metaphoric gestures and iconic, as categorized by McNiel. For the categories of the recorded gestures, metaphoric gestures appeared the most, 68.41% of all gestures, followed by 23.73% beat gestures, 4.76% iconic gestures, and 3.11% deictic gestures.
The Proceedings of Design & Systems Conference, 2004
Eng 血 eeh 皿 g supPor 仁 system fl) r medlanical design based on the tedhnology of Web Servioes 捻 p... more Eng 血 eeh 皿 g supPor 仁 system fl) r medlanical design based on the tedhnology of Web Servioes 捻 propesed and implemented as an 血 tegrated design system by CADICAMICAE 血 disthbuted envirx)mnellt ． Knowledge base system for mechanical design and numerical analysis can heip fo organize digital processes f (｝ r design by analys 追 ． Un 丗 ed 血 」 brmation of 丘 gure or ana 】 ysB and p買)oess 血g data by XML expwession can be used 血 each design process under physi 〔 nl testriCtions ． Web SelVices can proVide high end computing system for the eng 血 eeri 皿 g enVil ' enment for皿 ore mechanical engtiieer aiid show the possibdity ofintegratedmodel and eXperimental system as a basic system ofproduction knowledge．
Electronic and Photonic Packaging, Electrical Systems Design and Photonics, and Nanotechnology, 2002
In development of surface mount technology for Ball Grid Array (BGA) or flip-chip assemblies, it ... more In development of surface mount technology for Ball Grid Array (BGA) or flip-chip assemblies, it is important to reduce stress concentration in solder joints as it is immediately effective in improving the fatigue life of the assembly. Thus, the ability to predict and control the joint geometries is critical to obtaining robust and reliable designs of interconnects. Other than the issue of the bump shape, there are many problems concerning soldering such as bridging or self-alignment in which surface tension is definitely involved. This paper attempts to apply the FEM approach in solving these problems. Rigid-plastic FEM which is based on iteration for the velocity field in an incompressible viscous fluid is an approach to large deformation analysis. The flow stress is described by the viscosity of the fluid and the strain rate. By introducing an automatic mesh updating procedure, transient problems with free boundary can be treated. We applied this concept to the prediction of sold...
2015 IEEE Global Engineering Education Conference (EDUCON), 2015
Speech-based test is a one of the best ways to train students' level of proficiency. However,... more Speech-based test is a one of the best ways to train students' level of proficiency. However, in e-learning it is difficult for students to maintain their concentration when speaking to a monitor. We propose speech-based online test system employing a human-likely embodied graphical agent as a listener. The developed system consists of three modules, user recognition, agent's non-verbal behavior realization, and learning controller. The system is developed with HTML5, Javascript, and WebGL, so that it should be compatible with any modern Web-browsers. To investigate the effects of the use of the listener agent in speech-based test on learning efficacy and naturalness of speech, we will conduct a one-way between-subject design experiment with three groups. In experimental group, participants speech to a listening agent with listener behavior depicted on a monitor during a speech-based test. On the other hand, in first control group, participants take the test with the listening agent but the agent does not move while the participants' speech. In second control group, participants take the test without the listening agent. Then, the result of listening scores will be statistically tested.
Uploads
Papers by Hiroshi Sakuta