Suche
News
You are here
Gait Recognition based on Spatiotemporal Features of Human Motion |
| M. H. Khan, M. S. Farid, M. Grzegorzek |
| The proposed technique generates the dense trajectories from a video sequence using optical flow field and their motion information is encoded in local descriptors. A codebook based on Gaussian mixture model (GMM) is built from randomly selected one million motion descriptors to encode the features. The local descriptors are encoded using Fisher vector encoding and classified using Linear Support Vector Machine (SVM). Experimental results on five benchmark gait databases confirm the effectiveness of the proposed algorithm. |
| · Code (6 MB) · Sample Data (1.92 GB) |