SCA 18: Posters

Permanent URI for this collection

https://diglib.eg.org/handle/10.2312/2632600

Browse

Browsing SCA 18: Posters by Subject "Motion capture"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Capturing Floor Exercise from Multiple Panning-Zooming Cameras
    (The Eurographics Association, 2018) Kobayashi, D.; Yamamoto, Masanobu; Skouras, Melina
    By panning and zooming camera, a system of the multiple cameras can obtain wider range of common field of views. It means that an image based motion capture system can measure bodies in motion of wider range. To do so, a key idea is a camera calibration by matching the panned and zoomed image with a panoramic image of the background. We show an experiment of motion capture of a gymnastic athlete in floor exercise by the calibrated cameras.
  • Thumbnail Image
    Item
    Dilated Temporal Fully-Convolutional Network for Semantic Segmentation of Motion Capture Data
    (The Eurographics Association, 2018) Noshaba, Cheema; Hosseini, Somayeh; Sprenger, Janis; Herrmann, Erik; Du, Han; Fischer, Klaus; Slusallek, Philipp; Skouras, Melina
    Semantic segmentation of motion capture sequences plays a key part in many data-driven motion synthesis frameworks. It is a preprocessing step in which long recordings of motion capture sequences are partitioned into smaller segments. Afterwards, additional methods like statistical modeling can be applied to each group of structurally-similar segments to learn an abstract motion manifold. The segmentation task however often remains a manual task, which increases the effort and cost of generating large-scale motion databases. We therefore propose an automatic framework for semantic segmentation of motion capture data using a dilated temporal fully-convolutional network. Our model outperforms a state-of-the-art model in action segmentation, as well as three networks for sequence modeling. We further show our model is robust against high noisy training labels.