Abstract
The user quality of experience of current 3D Internet technology lags far behind the one of installed 3D applications. We believe that a broad uptake of ‘3D Media Internet’ depends on a significant improvement of the user quality of experience, which should not significantly differ from the quality offered by similar 3D services installed on the user’s device. The approach we propose combines compact shape representations with the use of the parallel computational performance of GPUs, multi-core CPUs and many-core CPUs.
