The number of stroke patients have been increased with social aging in recent years. A rehabilitation is necessary for many stroke patients because of aftereffect in brain function. A rehabilitation to speech disorder in higher brain dysfunction is one of important problems, and the Miyake's paired verbal associate learning test is known as one of the examination method. In the examination method, after an examiner reads out related word pairs and non-related word pairs, the examiner either bring one of the word pair, and an examinee answer the other of word pair. This examination is an approach to check the cognitive faculty by giving the auditory stimulation to the examiner. The Miyake's paired verbal associate learning test is a subjective approach that has some different inspection procedures by each examiner. In this paper, we propose the method with Android PC application in order to reduce the difference of procedure by each examiner in Miyake's learning test, and to perform personal test. And we think about the difference of the brain stimulation between the traditional method and our proposed method from the results of brain function measurement in fNIRS. The results show that the proposed method with PC application is efficient as new Miyake's learning test, since the brain blood flow of both method is a little different each other.

View full abstract

Lip movement recognition is one aspect of multimodal voice recognition systems. We focused on this technology and we developed a pronunciation training system using face recognition software in our laboratory. In this research, we carried out English pronunciation training using the pronunciation training system. Also, we created database for the pronunciation training in cooperation with English teachers in my school. The present research is an examination of English pronunciation training comparing the English teacher's lip movements with those of the English pronunciation learner. For the English pronunciation learner, improving their English pronunciation is a great incentive to study English. Therefore, we propose a method of evaluation of English pronunciation training.

View full abstract

This work is a summary of the 3D model production method of the garden lanterns using Autodesk Memento. Autodesk Memento is capture software which can make a high-definition 3D model from multiple photos. The 3D model is automatically created by uploading the photograph of the object which we photographed to a cloud service. The photographic method and the choice of the photographic material are important to 3D production using Autodesk Memento. We summarized the 3D data production method including the photography method of a good material photograph. This time we produced 3D data with a view to data preservation and utilization of cultural assets. We made a database of the stone lanterns installed the Zuiryuji Temple, Haccho street, and the tomb of the Kagahan feudal lord Maeda Toshinaga in Takaoka city, Toyama prefecture.

View full abstract