Juntao Ye

Technical Section: Fitting 3D garment models onto individual human models

Designing an elegant 3D virtual garment model for a 3D virtual human model is labor-intensive, because most existing garment models are custom-made for a specific human model and cannot be easily reused for other individuals. In this paper, we propose a novel method for fitting a given 3D garment model onto human models of various body shapes and poses. The fitting is accomplished by deforming the garment mesh to match the shapes of the human models by using a combination of the following: skeleton-driven volumetric deformation, garment-human shape similarity matching and evaluation, the constraints of garment-human contact, and garment-human ease allowance. Experiments show that our approach performs very well and has the potential to be used in the garment design industry.

A calibration process for tracking upper limb motion with inertial sensors

Real-time inertial tracking of human motion requires to attach inertial sensors to the major segments of a human body. Due to the intrinsic noises and drifts of the sensor data, as well as the non-rigidity of the human muscles and skins, the tracking could be unprecise, especially as time elapses. Therefore, an elaborate calibration process and a noise-insensitive tracking algorithm are needed. This paper presents an Extended Kalman Filter-based method for calibrating the relative orientation between human upper limb bones and the sensors attached to them. The calibration result, represented as a rotation matrix, is then used in the tracking of the arm motion with the knowledge of the physical segment kinematic model. Several experiments have been performed to validate the effectiveness of our method.

Geometric Flow Approach for Region-Based Image Segmentation

Geometric flows have been successfully used for surface modeling and designing, largely because they are inherently good at controlling geometric shape evolution. Variational image segmentation approaches, on the other hand, detect objects of interest by deforming certain given shapes. This motivates us to revisit the minimal partition problem for segmentation of images, and propose a new geometric flow-based formulation and solution to it. Our model intends to derive a mapping that will evolve given contours or piecewise-constant regions toward objects in the image. The mapping is approximated by B-spline basis functions, and the positions of the control points are to be determined. Starting with the energy functional based on intensity averaging, we derive a Euler-Lagrange equation and then a geometric evolution equation. The linearized system of equations is efficiently solved via a special matrix-vector multiplication technique. Furthermore, we extend the piecewise-constant model to a piecewise-smooth model which effectively handles images with intensity inhomogeneity.

Simulating Inextensible Cloth Using Impulses

Computer simulation of cloth is often plagued by springs being over‐stretched. The evaluation of impulses to prevent over‐stretching is explained step‐by‐step. Our impulse approach controls the length of springs by a nonlinear system and then a novel linearization of it into a symmetric positive definite system. The cloth/solid collision handling is integrated into the linearized system seamlessly. Some results based on this method are also presented.

A Unified Cloth Untangling Framework Through Discrete Collision Detection

We present an efficient and stable framework, called Unified Intersection Resolver (UIR), for cloth simulation systems where not only impending collisions but also pre‐existing penetrations often arise. These two types of collisions are handled in a unified manner, by detecting edge‐face intersections first and then forming penetration stencils to be resolved iteratively. A stencil is a quadruple of vertices and it reveals either a vertex‐face or an edge‐edge collision event happened. Each quadruple also implicitly defines a collision normal, through which the four stencil vertices can be relocated, so that the corresponding edge‐face intersection disappear. We deduce three different ways, i.e., from predefined surface orientation, from history data and from global intersection analysis, to determine the collision normals of these stencils robustly. Multiple stencils that constitute a penetration region are processed simultaneously to eliminate penetrations. Cloth trapped in pinched environmental objects can be handled easily within our framework. We highlight its robustness by a number of challenging experiments involving collisions.

Anisotropic Strain Limiting for Quadrilateral and Triangular Cloth Meshes

The cloth simulation systems often suffer from excessive extension on the polygonal mesh, so an additional strain‐limiting process is typically used as a remedy in the simulation pipeline. A cloth model can be discretized as either a quadrilateral mesh or a triangular mesh, and their strains are measured differently. The edge‐based strain‐limiting method for a quadrilateral mesh creates anisotropic behaviour by nature, as discretization usually aligns the edges along the warp and weft directions. We improve this anisotropic technique by replacing the traditionally used equality constraints with inequality ones in the mathematical optimization, and achieve faster convergence. For a triangular mesh, the state‐of‐the‐art technique measures and constrains the strains along the two principal (and constantly changing) directions in a triangle, resulting in an isotropic behaviour which prohibits shearing. Based on the framework of inequality‐constrained optimization, we propose a warp and weft strain‐limiting formulation. This anisotropic model is more appropriate for textile materials that do not exhibit isotropic strain behaviour.

Automatic skinning and animation of skeletal models

In this paper, we present an efficient yet easy-to-implement technique which performs automatic skinning and animation of skeletal models. At a pre-processing stage, a character model is firstly decomposed into a number of segments per bone basis, and each segment is then subdivided into several chunks. A convex cage is automatically created for each chunk. The skinning and animation of skeletal models is achieved via two steps. At the first step, by minimizing a sum of several energy terms, chunk cages are implicitly skinned to the skeleton and animated. These energies are carefully designed to prevent unnatural volume change and guarantee smooth deformation transition between adjacent cages. At the second step, the model mesh vertices, represented as the mean-value coordinates with reference to proper cage vertices, are updated via cage-based deformation technique. Our approach avoids the labor-intensive process of vertex weighting and cage generation. Given the motion of a skeleton, the character model can be animated automatically.

A reduced unconstrained system for the cloth dynamics solver

Modern direct solvers have been more and more widely used by computer graphics community for solving sparse linear systems, such as those that arise in cloth simulation. However, external constraints usually prevent a direct method from being used for cloth simulation due to the singularity of the constrained system. This paper makes two major contributions towards the re-introduction of direct methods for cloth dynamics solvers. The first one is an approach which eliminates all the constrained variables from the system so that we obtain a reduced, nonsingular and unconstrained system. As alternatives to the well-known MPCG algorithm, not only the original, unmodified PCG method, but also any direct method can be used to solve the reduced system at a lower cost. Our second contribution is a novel direct-iterative scheme applied for the reduced system, which is basically the conjugate gradient method using a special preconditioner. Specifically, we use the stiff part of the coefficient matrix, which we call the matrix core, as the preconditioner for the PCG. The inverse of this preconditioner is computed by any eligible direct solver. The direct-iterative method has proved to be more efficient than both direct and iterative methods. Our experiments show a factor of two speedup over direct methods when stiff springs are used, even greater improvements over the MPCG iterative method.

Transferring and fitting fixed-sized garments onto bodies of various dimensions and postures

Abstract Many virtual try-on systems involve transferring and fitting garments to bodies of various shapes and postures, with grade preservation. To achieve this goal, garments must be treated as elastic models and their deformation is controlled by the laws of dynamics. Moreover, a collision-free state must be maintained during the simulation, as well as in the final draping state. We present a complete pipeline that concentrates on solving two problems: (1) deforming the target body towards the reference body, and (2) simulating the garment with robust handling of not only impending but also pre-existing collisions. Our solution to the first problem is a skeleton-driven framework, which consists of a collection of techniques, including skeleton embedding, skeleton posture alignment and skeleton-driven mesh deformation. For skeleton posture alignment, we decouple the orientation of each joint into two components: swing and twist, and align them separately. Treating garment models as rigid, the deformed target ‘fits’ into the garments with as few penetrations as possible. When solving the second problem, the body/garment penetrations are untangled along with cloth simulation, so that a collision-free state can be achieved. After that, the deformed target body restores its original shape gradually, while the garments are physically simulated to maintain a collision-free state, until a final draping state is reached on the fully restored target body. Examples show that the proposed framework is effective for garment transfer and fitness evaluation, and can be potentially used in applications like online shopping or customization.

Modeling 3D human body with a smart vest

Abstract A novel human body reconstruction method based on measuring human body feature curves is proposed in this paper. Feature curves reflecting the cross section shape of human body are obtained by reconstructing the shapes of flexible tapes on an easy-to-operate smart vest that is dressed on the user. These tapes are specially marked to make the length of curves and spatial position on each marker perceptible, when the user rotates in front of a binocular stereo system. Taking these perceived geometric information as the input, precise feature curves are computed by solving a constrained optimization function. Individual human model is then generated by mapping the feature curves onto human shape space based on a multilayer neural network. Comparative experiments prove that our approach provides a low-costs way to reconstruct human models with competitive geometric precision (about 8mm point-to-point distance). Our approach can be potentially applied in many human body related product industries.