| Presentation | 2019-03-15 A Study on Teacher Behavior Classification based on ST-GCN for Traial Lesson Support System Sho Ooi, Yao Shunyu, Haruo Noma, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | A new teacher needs to give tuition from the first day at work. People who aim for the teacher can teach at teaching practices and a crammer. However, an classroom environment is different the actual school environment. In other words, the new teacher is difficult to train tuition in an actual environment. Therefore, we think that we need to develop a trial lesson system close to the actual environment for the new teacher. The trial lesson system has three functions, namely, to display the environment close to the actual school, to evaluate teacher behavior quantitatively, to reflect oneself an experienced video (give oneself a awareness). This study focuses to evaluate teacher behavior quantitatively. The quantitatively of teachers behavior needs to recognize the behavior and to analyze behavior transition. In this research, we defined ten kinds of teachers behavior, conducted behavior recognition using ST-GCN. As a result, we were able to recognize ten kinds of teachers behavior with 91 percent accuracy. In addition, we analyzed the behavior during the actual trial lesson and visualized the behavior transition. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Behavior classification / ST-GCN / Teacher behavior / Deep Learning |
| Paper # | ET2018-97 |
| Date of Issue | 2019-03-08 (ET) |
| Conference Information | |
| Committee | ET |
|---|---|
| Conference Date | 2019/3/15(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Naruto University of Education |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | IoT & Education/Learning Support, etc. |
| Chair | Yozo Miyadera(Tokyo Gakugei Univ.) |
| Vice Chair | Ryo Takaoka(Yamaguchi Univ.) |
| Secretary | Ryo Takaoka(Open Univ. of Japan) |
| Assistant | Megumi Kurayama(National Inst. of Tech., Hakodate College) / Masaru Okamoto(Hiroshima City Univ.) |
| Paper Information | |
| Registration To | Technical Committee on Educational Technology |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | A Study on Teacher Behavior Classification based on ST-GCN for Traial Lesson Support System |
| Sub Title (in English) | |
| Keyword(1) | Behavior classification |
| Keyword(2) | ST-GCN |
| Keyword(3) | Teacher behavior |
| Keyword(4) | Deep Learning |
| 1st Author's Name | Sho Ooi |
| 1st Author's Affiliation | Ritsumeikan University(Rits) |
| 2nd Author's Name | Yao Shunyu |
| 2nd Author's Affiliation | Ritsumeikan University(Rits) |
| 3rd Author's Name | Haruo Noma |
| 3rd Author's Affiliation | Ritsumeikan University(Rits) |
| Date | 2019-03-15 |
| Paper # | ET2018-97 |
| Volume (vol) | vol.118 |
| Number (no) | ET-510 |
| Page | pp.pp.59-62(ET), |
| #Pages | 4 |
| Date of Issue | 2019-03-08 (ET) |