講演抄録/キーワード |
講演名 |
2018-09-21 10:10
[ショートペーパー]Kinect RGB-D Hand Gesture Image Database for Deep Learning-Based Gesture Recognition ○Jiaqing Liu・Kotaro Furusawa・Seiju Tsujinaga(Ritsu)・Tomoko Tateyama(Hiroshima)・Yutaro Iwamoto・YenWei Chen(Ritsu) PRMU2018-58 IBISML2018-35 |
抄録 |
(和) |
(まだ登録されていません) |
(英) |
We present a new dataset named the MaHG-RGBD, a multi-angle view hand gesture for deep learning-based gesture recognition. This dataset has the following features compared with the existing ones. Firstly, both depth and color of segmented hand region images are recorded at the same time. Secondly, by using two camera views to capture hand gestures: our dataset allows researchers to combine information from multiple angles to overcome the ambiguity in gestures recognition. Thirdly, the various of different gestures classes: a total of 25 classes. Finally, we evaluate the recognition accuracy of 25 different hand gestures using deep learning methods to form a benchmark on this dataset. The MaHG-RGBD dataset is available at the link: http://www.iipl.is.ritsumei.ac.jp/MaHG-RGBD. |
キーワード |
(和) |
/ / / / / / / |
(英) |
R-GBD Dataset / hand gestures / multi-angle view / deep learning / / / / |
文献情報 |
信学技報, vol. 118, no. 219, PRMU2018-58, pp. 141-142, 2018年9月. |
資料番号 |
PRMU2018-58 |
発行日 |
2018-09-13 (PRMU, IBISML) |
ISSN |
Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
PDFダウンロード |
PRMU2018-58 IBISML2018-35 |