We propose Cloud Baking, a collaborative rendering architecture for dynamic Web3D scenes. In our architecture, the cloud renderer renders the scene with the global illumination (GI) information in a GI map; the web-based client renderer renders the scene with ambient lighting only and blends it with the GI map received from the cloud for the final scene. This approach allows the users to interact with the web scene and change the scene dynamically through the web interface end, yet move the computationally heavy tasks of global illumination computation to the cloud. A challenge we face is the interaction delay that causes the frames rendered on the cloud and the client to go out of sync. We propose to use 3D warping and a hole-filling algorithm designed for GI map to predict the late GI map. We show both quantitatively and visually the quality of the GI map produced using our method. Our prediction algorithm allows us to further reduce the frequency at which the GI map is computed and sent from the server, reducing both computational needs and bandwidth usage.
3D visualization of digital human becomes a key tool for the medical visualization, especially for medical education. Web3D technology has been commonly applied in this field. However, the quality of rendering is not expected for the medical purpose. Nowadays, global illumination (GI) map is an efficient tool for real-time lighting and shadow rendering. On the cloud baking server, a large number of rendered GI maps are generated under variety of configuration in the scene on the Web3D interface end. GI tree works on organizing these baked maps for reusing in the Web3D client. Meanwhile, it dispatches the existing baked maps directly in the case that the viewpoint appears in the duplicate positions in the Web3D client. This is the main stream solution of the cloud pre-rendering. However, it is a challenge to store and manage excessive rendered maps. In this paper, we propose a light-weight collaborative machine learning method for lighting and shadow rendering in medical applications. In this system, the conditional generative adversarial networks (GAN) works for generating the GI map instead of finding out the similar from number of stored maps, and we propose structure-aware 3D image warping method to improve the system performance. The experiments demonstrated that our proposed system not only guarantees the resolution of the GI map in the Web3D client, but also significantly reduces the rendering computational needs so as to improve the system performance.
1 August 2020 in Computational and Mathematical Methods in Medicine
With the increasing of depth and complexity of the convolutional neural network, parameter dimensionality and volume of computing have greatly restricted its applications. Based on the SqueezeNet network structure, this study introduces a block convolution and uses channel shuffle between blocks to alleviate the information jam. The method is aimed at reducing the dimensionality of parameters of in an original network structure and improving the efficiency of network operation. The verification performance of the ORL dataset shows that the classification accuracy and convergence efficiency are not reduced or even slightly improved when the network parameters are reduced, which supports the validity of block convolution in structure lightweight. Moreover, using a classic CIFAR-10 dataset, this network decreases parameter dimensionality while accelerating computational processing, with excellent convergence stability and efficiency when the network accuracy is only reduced by 1.3%.
27 September 2022 in Security and Communication Networks
The collaborative online 3D rendering system proposed in this paper ensures the quality of user experience and protects online rendering resources. In this system, the conditional generative adversarial network is used to calculate complex global illumination information instead of rendering them on cloud servers. The web front-end generates high-frequency direct lighting information in real-time and displays the final result which is a blend of front-end direct lighting information and back-end indirect lighting information. Experiments show that our proposed system can improve the rendering quality of the Web3D front-end, ensure Web-Cloud load balance, and protect rendering resources online.
1 October 2026 in IEEE Transactions on Mobile Computing