Graphics Reference
In-Depth Information
6.7 Conclusions
In this chapter, we presented a prototype for 3D videoconferencing with eye-gaze
correction. A virtual camera is placed behind the display, and its image is synthe-
sized with image-based rendering from multiple real cameras around the display,
using a modified plane sweeping algorithm. A GPU implementation yields real-time
performances at high visual quality.
Performance is even further increased, thanks to algorithmic complexity reduction
approaches. We presented a method to reduce the computational requirements by
adapting the depth range and reducing the number of hypothesis depth planes in the
search space, redistributing them to places with a high object density. This results in a
substantial processing performance increase without impeding on the visual quality
of the view synthesis.
References
1. Avidan S, Shashua A (1997) Novel view synthesis in tensor space. In: Proceedings of the
conference on computer vision and pattern recognition (CVPR), Washington. IEEE, pp 1034-
1040
2. Baker H, Bhatti NT, Tanguay D, Sobel I, Gelb D, Goss ME, Culbertson WB, Malzbender T
(2002) The coliseum immersive teleconferencing system. In: Proceedings of the international
workshop on immersive telepresence, Juan-les-Pins
3. Brown DC (1966) Decentering distortion of lenses. Photom Eng 32(3):444-462
4. Chen SE, Williams L (1993) View interpolation for image synthesis. In: Proceedings of the
conference on computer graphics and interactive techniques (SIGGRAPH), Anaheim, 1993.
ACM, pp 279-288
5. Chien S-Y, Yu S-H, Ding L-F, Huang Y-N, Chen L-G (2003) Efficient stereo video coding
system for immersive teleconference with two-stage hybrid disparity estimation algorithm. In:
Proceedings of the international conference on image processing (ICIP), pp 749-752
6. Criminisi A, Shotton J, Blake A, Torr PHS (2003) Gaze manipulation for one-to-one tele-
conferencing. In: Proceedings of the ninth IEEE international conference on computer vision
(ICCV), Washington. IEEE, pp 191-198
7. Dumont M, Maesen S, Rogmans S, Bekaert P (2008) A prototype for practical eye-gaze cor-
rected video chat on graphics hardware. In: Proceedings of the international conference on
signal processing and multimedia applications (SIGMAP), Porto, July 2008. INSTICC, pp
236-243
8. Maarten D, Sammy R, Steven M, Philippe B (2009) Optimized two-party video chat with
restored eye contact using graphics hardware. Springer Comm Comput Inf Sci 48(11):358-
372
9. Geys I, Gool LV (2004) Extended view interpolation by parallel use of the GPU and the CPU.
In: Proceedings of the society of photo-optical instrumentation engineers (SPIE) conference:
videometrics VIII, vol 5665, pp 96-107
10. Geys I, Koninckx TP, Gool LV (2004) Fast interpolated cameras by combining a GPU based
plane sweep with a max-flow regularisation algorithm. In: Proceedings of the 2nd international
symposium on 3D data processing, visualization, and transmission (3DPVT), Washington.
IEEE, pp 534-541
11. Goorts P, Rogmans S, Bekaert P (2007) Optimal data distribution for versatile finite impulse
response filtering on next-generation graphics hardware using CUDA. In: Proceedings of the
