研究課題/領域番号 |
17K02964
|
研究機関 | 富山高等専門学校 |
研究代表者 |
COOPER T・D 富山高等専門学校, 一般教養科, 准教授 (70442449)
|
研究分担者 |
的場 隆一 富山高等専門学校, その他部局等, 准教授 (30592323)
塚田 章 富山高等専門学校, その他部局等, 教授 (40236849)
五味 伸之 福井工業高等専門学校, 機械工学科, 講師 (80600634)
成瀬 喜則 富山大学, 大学院教職実践開発研究科, 教授 (00249773)
|
研究期間 (年度) |
2017-04-01 – 2020-03-31
|
キーワード | Nonverbal communication / Facial expression / Speech wave analysis |
研究実績の概要 |
At present, we are developing three components of the system concurrently. The system is comprised of the following: analysis of facial expression (including facial recognition); analysis of voice quality (individual sounds, intonation, and word/sentence stress); and gestural recognition. This year we presented our findings at 3 international conferences in Holland, Hong Kong and Cambodia. Facial expression was improved gaining higher accuracy from a smaller catalog of images. An algorithm was created to align with the ratings of a human judge. Finally, we have started investigating how A.I. (artificial intelligence) can help improve our research.
|
現在までの達成度 (区分) |
現在までの達成度 (区分)
3: やや遅れている
理由
There has been a slight delay due to various personnel changes, resulting in being unable to implement the software into the curricula.
|
今後の研究の推進方策 |
We are currently improving our system in the three following areas: the facial expression analysis will be improved with AI, requiring less personal data, as well as the facial recognition component; gestural analysis will be improved along the same method; speech audio wave analysis will have its evaluation criterion expanded. The increase of evaluation criterion in the rubric will increase the accuracy of the system.
|
次年度使用額が生じた理由 |
While we were successful in having three presentations at international conferences, two of the co-researchers had their papers not accepted. We will revise and improve those findings with the incurring amount.
|