A5101986644- Computer Graphics and Visualization Techniques
- Data Visualization and Analytics
- Generative Adversarial Networks and Image Synthesis
- 3D Shape Modeling and Analysis
- Image Enhancement Techniques
- Advanced Vision and Imaging
- Data Analysis with R
- Time Series Analysis and Forecasting
- Model Reduction and Neural Networks
- Visual Attention and Saliency Detection
- Human Pose and Action Recognition
- Hand Gesture Recognition Systems
- Wind and Air Flow Studies
- Advanced Numerical Analysis Techniques
- Teleoperation and Haptic Systems
- Semantic Web and Ontologies
- Tactile and Sensory Interactions
- Smart Agriculture and AI
- Advanced Clustering Algorithms Research
- Gait Recognition and Analysis
- Advanced Image Fusion Techniques
- Wood and Agarwood Research
- Fluid Dynamics and Turbulent Flows
- Meteorological Phenomena and Simulations
- Advanced Numerical Methods in Computational Mathematics
Zhejiang University of Technology
2019-2024
Walt Disney (Switzerland)
2023-2024
ETH Zurich
2019-2023
Tokushima University
2019
Walt Disney (United States)
2019
The goal of natural image matting is the estimation opacities a user-defined foreground object that essential in creating realistic composite imagery. Natural challenging process due to high number unknowns mathematical modeling problem, namely as well and background layer colors, while original serves single observation. In this paper, we propose colors through use deep neural networks prior opacity estimation. color better match for capabilities networks, availability these substantially...
Identifying causality behind complex systems plays a significant role in different domains, such as decision-making, policy implementations, and management recommendations. However, existing studies on temporal event sequence data mainly focus individual causal discovery, which is incapable of capturing combined causality. To address the gap discovery data, eliminating recruiting principles are defined to balance effectiveness controllability cause combinations. We also leverage Granger...
Abstract Fluid control often uses optimization of forces that are added to a simulation at each time step, such the final animation matches single or multiple target density keyframes provided by an artist. The problem is strongly under‐constrained with high‐dimensional parameter space, and finding optimal solutions challenging, especially for higher resolution simulations. In this paper, we propose two novel ideas jointly tackle lack constraints high dimensionality space. We first consider...
Humans encounter a vast array of sensory stimuli in their everyday lives. However, many visualization techniques primarily utilize visual feedback, which may disregard certain intricate details. Relying on single channel overlook complex layouts. how haptic force feedback can be used to assist remained under-explored. In this work, we initially conducted literature review identify potential problems the large datasets and engaged discussions with domain experts explore collision...
Abstract Creative processes of artists often start with hand‐drawn sketches illustrating an object. Pre‐visualizing these keyframes is especially challenging when applied to volumetric materials such as smoke. The authored 3D density volumes must capture realistic flow details and turbulent structures, which highly non‐trivial remains a manual time‐consuming process. We therefore present method compute smoke field directly from 2D artist sketches, bridging the gap between early‐stage...
The wide application of personal biometric information such as face, fingerprint, iris, and voiceprint has simultaneously created many new ethical legal issues, including the fraudulent use biometrics. A non-human system is demanded an alternative, which features no human private can be replaced or renewed from time to time. main objective this study identify wood leaf patterns verify their identities by building respective datasets. On basis, a plant feature-based recognition authentication...
An accurate assessment of physical transport requires high-resolution and high-quality velocity information. In satellite-based wind retrievals, the accuracy is impaired due to noise while maximal observable resolution bounded by sensors. The reconstruction a continuous field important assess characteristics it very challenging. A major difficulty ambiguity, since lack visible clouds results in missing information multiple fields will explain same sparse observations. It is, therefore,...
While controlling simulated gaseous volumes remains an ongoing battle when seeking realism in computer graphics, creating appealing characters entirely out of these simulations brought this challenge to new level Pixar's film Elemental. For fire characters, like the protagonist "Ember", their faces and bodies needed look move real fire, but not be so frenetic as distract from acting emotion performances. Neural Style Transfer emerged a key technique achieving that met criteria. By using more...
Hand Gesture Recognition plays an important role in noncontact human computer interaction. The traditional based on vision has specific requirements for the target location and equipment. We employ monocular camera to build a system which can detect hand any position of screen recognize gesture. Single Shot MultiBox Detector (SSD) locates Convolutional Neural Network (CNN) makes classification gestures. gesture recognition performance proposed technique is 95.05% Massey University dataset,...
Abstract An ongoing challenge in fluid animation is the faithful preservation of vortical details, which impacts visual depiction flows. We propose Impulse Particle‐In‐Cell (IPIC) method, a novel extension popular Affine (APIC) method that makes use impulse gauge formulation equations. Our approach performs coupled advection‐stretching during particle‐based advection to better preserve circulation and details. The associated algorithmic changes are simple straightforward implement, our...
Abstract Controlling fluid simulations is notoriously difficult due to its high computational cost and the fact that user control inputs can cause unphysical motion. We present an interactive method for deformation‐based control. Our aims at balancing direct deformations of fields preservation physical characteristics. train convolutional neural networks with physics‐inspired loss functions together a differentiable simulator, provide efficient workflow flow manipulations test time....