Editorial Issue 34.5

Abstract

This issue contains 12 papers. Seven papers have been selected from the CASA 2022 Aninex workshop by the program committee chaired by Professor Jian Chang from Bournemouth University. These papers have been extensively revised and then reviewed by the CAVW editorial team. The last five papers are regular papers.

In the first paper, Wenshu Zhang et al. present Struct2Hair, a novel single-viewed hair modeling approach by extracting hair shape descriptor (HSD). The HSD is defined as the fundamental structure-aware feature, which is a combination of critical shapes in a hairstyle. A complete dataset of critical hair shapes is constructed from a known database of 3D hair models.

In the second paper, Yanrui Xu et al. propose a novel boundary-distance based adaptive method for SPH fluid simulation. The signed-distance field constructed concerning the coupling boundary is introduced to determine particle resolution in different spatial positions. The resolution is maximal within a specific distance to the boundary and decreases smoothly as the distance increases until a threshold is reached. The sizes of the particles are then adjusted toward the resolution via splitting and merging. Additionally, a wake flow preservation mechanism is introduced to keep the particle resolution at a high level for a period of time after a particle flows through the boundary object to prevent the loss of flow details.

In the third paper, Tingting Li et al. propose a point cloud synthesis method based on stochastic differential equations (SDEs). They view the point cloud generation process as smoothly transforming from a known prior distribution toward the high-likelihood shape by point-level denoising. They introduce a conditional corrector sampler to improve the quality of point clouds. They additionally prove that their approach can be trained in an auto-encoding fashion and reconstruct point cloud faithfully. Furthermore, their model can be extended on a downstream application of point clouds completion. Experimental results demonstrate the effectiveness and efficiency of their method.

In the fourth paper, Shuqing Yu et al. present a multiscale framework with visual field analysis branch to improve estimation accuracy. The model is based on the feature pyramids and predicts vision field to help gaze estimation. In particular, the authors analyze the effect of the multiscale component and the visual field branch on challenging benchmark datasets: MPIIGaze and EYEDIAP. Based on these studies, their proposed PerimetryNet significantly outperforms state-of-the-art methods. In addition, the multiscale mechanism and visual field branch can be easily applied to existing network architecture for gaze estimation.

The fifth paper by Junheng Fang et al. focuses on the emergence of position-based simulation approaches that has quickly developed a group of new topics in the computer graphics community. These approaches are popular due to their advantages, including computational efficiency, controllability, stability, and robustness for different scenarios, while they also have some weaknesses. In this survey, the authors introduce the concept of the baseline position-based dynamics (PBD) method and review the improvements and applications of PBD since 2018, including extensions for different materials and integrations with other techniques.

In the sixth paper, Xiaokun Wang et al. propose an implicit smoothed particle hydrodynamics fluid-elastic coupling approach to reduce the instability issue for fluid–fluid, fluid-elastic, and elastic–elastic coupling circumstances. By deriving the relationship between the universal pressure field with the incompressible attribute of the fluid, the authors apply the number density scheme to solve the pressure Poisson equation for both fluid and elastic material to avoid the density error for multimaterial coupling and conserve the nonpenetration condition for elastic objects interacting with fluid particles. Experiments show that their method can effectively handle the multiphase fluids simulation with elastic objects under various physical properties.

In the seventh paper, Hui Liang et al. propose a semantic-based scene generation method for digital shadow puppet performance scene. According to this method, the key information is extracted from the descriptive text using the Chinese text segmentation technology. Meanwhile, the authors generate semantic scene graphs and search the corresponding shadow puppet models in the model library to construct the virtual scenes of digital shadow puppet performance.

In the first regular paper, Marco Cirelli et al. propose by means of a virtual prototype and multibody dynamics simulation, the physical feasibility of a method for the ascending of rock blocks for building the Egyptian pyramids. From historical and archeological bases, this investigation presents the fundamental functional features of the virtual model components for the ascending of the stones. Furthermore, the methodological details for the model setup, as well as the discussion on the stone ascending movements, are herein addressed. The main results obtained from the simulation include the evaluation of the advancement-per-cycle of the conjectured ascending device and the corresponding required driving forces.

In the second regular paper, Ye Zhang et al. report on a meta-learning based method for finger vein recognition to overcome the problem of low recognition accuracy caused by the small number and variety of finger vein samples as well as fuzzy vein lines. The method is based on meta-learning, incorporating multiscale features, and using the idea of residual networks to join meta-learning to improve the recognition accuracy of finger vein images with few samples; to further improve its recognition ability, a differential map is constructed in the form of a differential between the finger vein image of singular value decomposition and finger vein image.

The third regular paper by TaeYoung Kim et al. presents a method of directly measuring the reflection coefficient of a surface, which is an acoustic characteristic in the real environment. Because expensive optimization-based studies that mainly aim to reproduce recorded sounds indirectly estimate acoustic materials, new estimates are required whenever the actual environment changes. Their approach utilizes the method of the acoustics field to enable anyone to easily and directly measure the reflection coefficient of a real environment and generate sound in a virtual environment. The authors obtain the impulse response (IR) for the target surface, separate the direct sound and the reflected sound, and calculate the reflection coefficient for each surface.

In the fourth regular paper, Sandhya Rani Sahoo et al. propose a customized deep convolutional neural network (CNN) architecture that has been designed to discriminate between benign and malignant lesions. The model is designed carefully with lesser convolution layers, fewer filters, and parameters to achieve better classification performance compared with pretrained models and ensures state-of-the-art performance. The proposed model is composed of nine trainable layers: eight convolution layers and one fully connected layer. The suggested framework is extensively evaluated on the benchmark ISIC 2016 challenge dataset. The effect of different input transformations over the dataset has been studied. For fair comparison, standard deep learning models have been used for lesion classification using transfer learning approach. Results show that class balancing with external images improves classification performance.

The last regular paper by V. Jothi Prakash et al. propose a hypothetical meta-stack framework to understand the various components in the realm of metaverse and then provide wide-ranging insights on the most recent development in metaverse realm in the context of cutting-edge technologies, security vulnerabilities, and preventive measures specific to the metaverse and the research challenges pertaining to metaverse.