Coarse-to-Fine Network-Based Intra Prediction in Versatile Video Coding.

After the development of the Versatile Video Coding (VVC) standard, research on neural network-based video coding technologies continues as a potential approach for future video coding standards. Particularly, neural network-based intra prediction is receiving attention as a solution to mitigate the limitations of traditional intra prediction performance in intricate images with limited spatial redundancy. This study presents an intra prediction method based on coarse-to-fine networks that employ both convolutional neural networks and fully connected layers to enhance VVC intra prediction performance. The coarse networks are designed to adjust the influence on prediction performance depending on the positions and conditions of reference samples. Moreover, the fine networks generate refined prediction samples by considering continuity with adjacent reference samples and facilitate prediction through upscaling at a block size unsupported by the coarse networks. The proposed networks are integrated into the VVC test model (VTM) as an additional intra prediction mode to evaluate the coding performance. The experimental results show that our coarse-to-fine network architecture provides an average gain of 1.31% Bjøntegaard delta-rate (BD-rate) saving for the luma component compared with VTM 11.0 and an average of 0.47% BD-rate saving compared with the previous related work.

Incomplete Region Estimation and Restoration of 3D Point Cloud Human Face Datasets.

Owing to imperfect scans, occlusions, low reflectance of the scanned surface, and packet loss, there may be several incomplete regions in the 3D point cloud dataset. These missing regions can degrade the performance of recognition, classification, segmentation, or upsampling methods in point cloud datasets. In this study, we propose a new approach to estimate the incomplete regions of 3D point cloud human face datasets using the masking method. First, we perform some preprocessing on the input point cloud, such as rotation in the left and right angles. Then, we project the preprocessed point cloud onto a 2D surface and generate masks. Finally, we interpolate the 2D projection and the mask to produce the estimated point cloud. We also designed a deep learning model to restore the estimated point cloud to improve its quality. We use chamfer distance (CD) and hausdorff distance (HD) to evaluate the proposed method on our own human face and large-scale facial model (LSFM) datasets. The proposed method achieves an average CD and HD results of 1.30 and 21.46 for our own and 1.35 and 9.08 for the LSFM datasets, respectively. The proposed method shows better results than the existing methods.