The KS20 VisLab Multi-View Kinect skeleton dataset is a set of multi-view Kinect skeleton (KS) data sequences collected from 20 walking subjects using Kinect V.2., in the context of long-term person re-identification using biometrics. Multiple walking sequences along five different directions i.e., Left lateral (LL at ~0o), Left diagonal (LD at ~ 30o), Frontal (F at ~ 90o), Right diagonal (RD at ~ 130o) and Right lateral (RL at ~180o) were collected. Altogether we have 300 skeleton image sequences comprising 20 subjects (3 video sequences per person in a particular viewpoint) in the aforementioned directions.
Keywords: Video Surveillance, Person Re-identification, Soft-biometrics, Gait.
1) Context-Aware Person Re-identification in the Wild via fusion of Gait and Anthropometric features, A. Nambiar, A. Bernardino, J. C. Nascimento and A. Fred, B-WILD Workshop at 12th IEEE International Conference on Automatic Face & Gesture Recognition (FG). Washington DC, USA, 30 May – 3 June 2017.
2) Towards view-point invariant Person Re-identification via fusion of Anthropometric and Gait Features from Kinect measurements, A. Nambiar, A. Bernardino, J. C. Nascimento, A. Fred, International Conference on Computer Vision Theory and Applications (VISAPP), Porto, Portugal, Feb. 2017.
Access to the Vislab Multi-view KS20 dataset is available upon request. Maintained by Athira Nambiar (anambiar at isr dot tecnico dot ulisboa dot pt). Contact me if you are interested in this dataset.
Copyright © 2017. Proprietary and Confidential to Vislab, Institute for Systems and Robotics (ISR/IST Lisboa).