Presenter Localization for Augmented Virtual Presentation

Combined Reality (augmented virtual reality, AVR) is gaining much more popularity in sports video comment, especially in the soccer match video. Compared with the ways of traditional presenting, Augmented Virtual Presentation (AVP) can provide audience a kind of immersive feeling. One of the key technologies of AVP is player localization. In this paper, we propose a new method of player localization in AVP system. We first detect the commented players with an improved SSD model, and then select a proper frame according to importance evaluation of the frames. Finally, we find the optimal position where we want to place the presenter. The extensive experiments demonstrate the effectiveness of our method and advantages over the traditional ones.

[1]  Andrew Hogue,et al.  A real-time reconstructed 3D environment augmented with virtual objects rendered with correct occlusion , 2014, 2014 IEEE Games Media Entertainment.

[2]  Wei Liu,et al.  SSD: Single Shot MultiBox Detector , 2015, ECCV.

[3]  Teijo Lehtonen,et al.  Extended panorama tracking algorithm for augmenting virtual 3D objects in outdoor environments , 2016, 2016 22nd International Conference on Virtual System & Multimedia (VSMM).

[4]  Byeong-Seob Ko,et al.  Sports highlights generation bas ed on acoustic events detection: A rugby case study , 2015, 2015 IEEE International Conference on Consumer Electronics (ICCE).

[5]  Jian Sun,et al.  Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition , 2014, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[6]  Ali Farhadi,et al.  You Only Look Once: Unified, Real-Time Object Detection , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[7]  Jan-Michael Frahm,et al.  Development of vision-aided navigation for a wearable outdoor augmented reality system , 2014, 2014 IEEE/ION Position, Location and Navigation Symposium - PLANS 2014.

[8]  Iasonas Kokkinos,et al.  Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs , 2014, ICLR.

[9]  Trevor Darrell,et al.  Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation , 2013, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[10]  Jeongkyu Lee,et al.  Event detection on large social media using temporal analysis , 2017, 2017 IEEE 7th Annual Computing and Communication Workshop and Conference (CCWC).