Title | ||
---|---|---|
Gait metric learning siamese network exploiting dual of spatio-temporal 3D-CNN intra and LSTM based inter gait-cycle-segment features. |
Abstract | ||
---|---|---|
•We have proposed a new representation from gait video frames, stereo silhouettes maps.•Each gait cycle is broken into a few gait cycle segments (GCS) derived from human gait biomechanics.•Spatio-temporal features are extracted for learning intra GCS relationship using pre-trained 3-D CNN.•Over the intra GCS features, LSTM is used to learn long and short term inter GCS relationship.•The gait metric is learned by training in a Siamese framework using triplet loss function, with dynamic adaptive margin. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1016/j.patrec.2019.07.008 | Pattern Recognition Letters |
Keywords | Field | DocType |
Gait biometrics,Deep learning,3-D Convolutional neural network,LSTM,Siamese | Computer vision,Pattern recognition,Gait,Silhouette,Gait cycle,Transfer of learning,Artificial intelligence,Concatenation,Biometrics,Mathematics,Binary number | Journal |
Volume | ISSN | Citations |
125 | 0167-8655 | 0 |
PageRank | References | Authors |
0.34 | 0 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Daksh Thapar | 1 | 2 | 3.54 |
Gaurav Jaswal | 2 | 22 | 6.23 |
Aditya Nigam | 3 | 154 | 28.82 |
Chetan Arora | 4 | 296 | 29.51 |