Abstract
| Original language | English |
|---|---|
| Pages (from-to) | 731-735 |
| Journal | IEEE Signal Processing Letters |
| DOIs | |
| Publication status | Published - 2017 |
Fingerprint
Cite this
}
SkeletonNet: Mining Deep Part Features for 3-D Action Recognition. / Ke, Qiuhong; An, Senjian; Bennamoun, Mohammed; Sohel, Ferdous; Boussaid, Farid.
In: IEEE Signal Processing Letters, 2017, p. 731-735.Research output: Contribution to journal › Article
TY - JOUR
T1 - SkeletonNet: Mining Deep Part Features for 3-D Action Recognition
AU - Ke, Qiuhong
AU - An, Senjian
AU - Bennamoun, Mohammed
AU - Sohel, Ferdous
AU - Boussaid, Farid
PY - 2017
Y1 - 2017
N2 - This letter presents SkeletonNet, a deep learning framework for skeleton-based 3-D action recognition. Given a skeleton sequence, the spatial structure of the skeleton joints in each frame and the temporal information between multiple frames are two important factors for action recognition. We first extract body-part-based features from each frame of the skeleton sequence. Compared to the original coordinates of the skeleton joints, the proposed features are translation, rotation, and scale invariant. To learn robust temporal information, instead of treating the features of all frames as a time series, we transform the features into images and feed them to the proposed deep learning network, which contains two parts: one to extract general features from the input images, while the other to generate a discriminative and compact representation for action recognition. The proposed method is tested on the SBU kinect interaction dataset, the CMU dataset, and the large-scale NTU RGB+D dataset and achieves state-of-the-art performance.
AB - This letter presents SkeletonNet, a deep learning framework for skeleton-based 3-D action recognition. Given a skeleton sequence, the spatial structure of the skeleton joints in each frame and the temporal information between multiple frames are two important factors for action recognition. We first extract body-part-based features from each frame of the skeleton sequence. Compared to the original coordinates of the skeleton joints, the proposed features are translation, rotation, and scale invariant. To learn robust temporal information, instead of treating the features of all frames as a time series, we transform the features into images and feed them to the proposed deep learning network, which contains two parts: one to extract general features from the input images, while the other to generate a discriminative and compact representation for action recognition. The proposed method is tested on the SBU kinect interaction dataset, the CMU dataset, and the large-scale NTU RGB+D dataset and achieves state-of-the-art performance.
U2 - 10.1109/LSP.2017.2690339
DO - 10.1109/LSP.2017.2690339
M3 - Article
SP - 731
EP - 735
JO - IEEE Signal Processing Letters
JF - IEEE Signal Processing Letters
SN - 1070-9908
ER -