Collision detection method, system, equipment, medium and product in clothing simulation

Abstract

The present application discloses a collision detection method, system, device, medium and product in clothing simulation, which relates to the fields of computer vision and graphics processing. It is the first time that a deep learning network is used for collision detection in clothing simulation. By training an MLP-based skin weight field prediction network and an implicit geometric representation deep learning network, accurate collision detection in clothing simulation is achieved using the MLP-based skin weight field prediction network and the implicit geometric representation deep learning network. Compared with the computationally expensive method of generating a discrete SDF or occupancy field composed of spatial voxels for the bounding box where the human body is located and performing detailed collision detection through spatial hashing, the present application can significantly reduce the computational overhead while accurately detecting and resolving collisions between fabric and the human body.

Description

Technical Field

[0001] The present application relates to the field of computer vision and graphics processing technology, and in particular to a collision detection method, system, device, medium and product in clothing simulation. Background Art

[0002] Collision detection (i.e., collision handling) in clothing simulation is used to detect and resolve collisions between fabric and the human body. If collisions cannot be determined and resolved during the simulation, severe wear and tear can occur in physical simulations of fabric and other objects, defeating the purpose of using computers to simulate the behavior of real-world objects. Therefore, collision detection in clothing simulation is crucial.

[0003] Collision detection (i.e., collision handling) in clothing simulation typically involves two steps: collision detection, which determines whether a collision has occurred, and collision resolution, which resolves the collision point. Collision handling typically utilizes gradient information.

[0004] Currently, cloth simulation algorithms use simple spheres or capsules to fit the geometry occupied by the human body. Clothing simulation typically involves two rounds of collision detection. First, a coarse-grained collision detection is performed using the capsule or sphere-fitted human body to filter out most points. Then, a discrete spatial dimension (SDF) or occupancy field (OF) is generated from the bounding box containing the human body. This is then used to perform a fine-grained collision detection using spatial hashing. However, this approach suffers from the fact that the effectiveness of the generated collision field processing is dependent on the accuracy of the discretization. The finer the division, the better the geometric fit for the human body, but the greater the computational overhead of constructing the collision field. Therefore, current collision detection algorithms for clothing simulation cannot achieve both accurate collision detection and resolution between cloth and the human body while significantly reducing computational overhead. Summary of the Invention

[0005] The purpose of this application is to provide a collision detection method, system, device, medium and product in clothing simulation, which can accurately detect and resolve collisions between fabric and human body while significantly reducing computational overhead.

[0006] To achieve the above objectives, this application provides the following solutions:

[0007] In a first aspect, the present application provides a collision detection method in clothing simulation, the collision detection method in clothing simulation comprising:

[0008] Acquire multiple action sequences provided by a subject with specified morphological parameters; each of the action sequences corresponds to an action; each of the action sequences includes posture parameters of multiple frames;

[0009] For each frame's posture parameter, input the posture parameter into the SMPL model to obtain the SMPL model under the posture parameter;

[0010] Based on the SMPL model under each posture parameter, obtain the sampling space corresponding to each frame;

[0011] For each frame, a plurality of different position points are sampled in the sampling space, wherein half of the position points are obtained by spatially uniform sampling and the other half of the position points are obtained by dense Poisson disk sampling on the human body surface;

[0012] A training dataset is constructed using all the position points sampled in the sampling space corresponding to each frame, and whether each position point is inside or outside the human body's occupancy field; 1 indicates that the position point is inside the human body's occupancy field, and 0 indicates that the position point is outside the human body's occupancy field;

[0013] Construct an MLP-based skin weight field prediction network and an implicit geometry representation deep learning network, including an MLP network for the skin weight field and an implicit geometry representation occupancy network for the human body under T-Pose, to obtain a collision detection model.

[0014] The collision detection model is trained using the training data set to obtain a trained collision detection model; an MLP network of a skin weight field in the trained collision detection model is used to predict skin weights for an input position point to be detected using a differentiable forward skinning network, and a T-Pose space corresponding point is obtained based on the skin weights in combination with a Broyden method; an occupancy network of an implicit geometric representation of the human body under T-Pose in the trained collision detection model is used to determine whether the position point to be detected is inside or outside the human body's occupancy field based on the input T-Pose space corresponding point;

[0015] Get the position of the vertex of the clothing to be detected;

[0016] The trained collision detection model is used to determine whether the vertex of the clothing to be detected is inside or outside the occupation field of the human body. If the vertex of the clothing to be detected is outside the occupation field of the human body, no collision occurs between the cloth and the human body. If the vertex of the clothing to be detected is inside the occupation field of the human body, a collision occurs between the cloth and the human body. The collided vertex is moved in the opposite direction of the gradient to resolve the collision.

[0017] Optionally, the action sequence is obtained through the DFaust subset of the AMASS dataset.

[0018] Optionally, the MLP network of the skin weight field includes a 4*128 MLP and a Softmax module, the 4*128 MLP is followed by a softmax module, and softplus is used as the activation function.

[0019] Optionally, the implicit geometric representation of the human body under T-Pose occupies a network including an 8*256 MLP and a Sigmoid module. The 8*256 MLP is followed by the Sigmoid module, and softplus is used as the activation function. The 8*256 MLP will make a skip connection at the 5th layer.

[0020] Optionally, the gradient is obtained through the trained collision detection model.

[0021] In a second aspect, the present application provides a collision detection system in clothing simulation, the collision detection system in clothing simulation comprising:

[0022] An action sequence acquisition module is used to acquire multiple action sequences provided by a subject with specified morphological parameters; each action sequence corresponds to an action; and each action sequence includes posture parameters of multiple frames;

[0023] An SMPL model acquisition module under each posture parameter is used to input the posture parameter into the SMPL model for each frame posture parameter to obtain the SMPL model under the posture parameter;

[0024] The sampling space acquisition module corresponding to each frame is used to obtain the sampling space corresponding to each frame based on the SMPL model under each posture parameter;

[0025] A different position point sampling module is used to sample multiple different position points in the sampling space corresponding to each frame, where half of the position points are obtained by spatially uniform sampling and the other half are obtained by dense Poisson disk sampling on the human body surface;

[0026] A training data set construction module is used to construct a training data set using all the position points sampled in the sampling space corresponding to each frame, and whether each position point is inside or outside the human body's occupation field; where 1 indicates that the position point is inside the human body's occupation field, and 0 indicates that the position point is outside the human body's occupation field;

[0027] The collision detection model construction module is used to build an MLP-based skin weight field prediction network and an implicit geometric representation deep learning network, including an MLP network for the skin weight field and an implicit geometric representation occupancy network for the human body under T-Pose, to obtain a collision detection model.

[0028] A collision detection model training module is configured to train the collision detection model using the training data set to obtain a trained collision detection model; an MLP network of the skin weight field in the trained collision detection model is configured to predict skin weights for input positions to be detected using a differentiable forward skinning network, and to obtain corresponding points in T-Pose space based on the skin weights in combination with the Broyden method; an occupancy network of the implicit geometric representation of the human body under T-Pose in the trained collision detection model is configured to determine whether the position to be detected is inside or outside the human body's occupancy field based on the input corresponding points in T-Pose space;

[0029] A clothing vertex position acquisition module is used to obtain the position of the vertex of the clothing to be detected;

[0030] The collision detection module is used to use the trained collision detection model to determine whether the vertex of the clothing to be detected is inside or outside the occupied field of the human body. If the vertex of the clothing to be detected is outside the occupied field of the human body, then no collision occurs between the cloth and the human body. If the vertex of the clothing to be detected is inside the occupied field of the human body, then a collision occurs between the cloth and the human body. The collided vertex is moved in the opposite direction of the gradient to resolve the collision.

[0031] Optionally, the gradient is obtained through the trained collision detection model.

[0032] In a third aspect, the present application provides a computer device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the collision detection method in the clothing simulation described above.

[0033] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any one of the above-mentioned collision detection methods in clothing simulation.

[0034] In a fifth aspect, the present application provides a computer program product, comprising a computer program, which, when executed by a processor, implements any one of the above-mentioned collision detection methods in clothing simulation.

[0035] According to the specific embodiments provided in this application, this application has the following technical effects:

[0036] The present application provides a collision detection method, system, device, medium and product in clothing simulation, which uses a deep learning network for collision detection in clothing simulation for the first time. By training an MLP-based skin weight field prediction network and an implicit geometric representation deep learning network, accurate collision detection in clothing simulation is achieved using the MLP-based skin weight field prediction network and the implicit geometric representation deep learning network. Compared with the computationally expensive method of generating a discrete SDF or occupancy field composed of spatial voxels for the bounding box where the human body is located and performing detailed collision detection through spatial hashing, the present application can significantly reduce the computational overhead while accurately detecting and resolving collisions between fabric and the human body. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0038] Figure 1 A schematic flow chart of a collision detection method in clothing simulation provided in one embodiment of the present application;

[0039] Figure 2 Flowchart used for collision detection in this application;

[0040] Figure 3 This is a detailed structural diagram of the MLP network of the skin weight field in this application;

[0041] Figure 4 This is the specific structure diagram of the static shape field MLP network of T-pose in this application;

[0042] Figure 5 Additional explanation diagram for gradient backpropagation in the training algorithm provided by this application;

[0043] Figure 6 Flowchart of the algorithm structure provided for this application;

[0044] Figure 7 This is a diagram illustrating a module for automatically calculating output gradients using a self-differentiation framework in the network structure provided by this application;

[0045] Figure 8 An overview of the clothing simulation collision processing algorithm based on implicit human probability field provided by this application;

[0046] Figure 9 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0048] The purpose of this application is to provide a collision detection method, system, device, medium and product in clothing simulation, which can accurately detect and resolve collisions between fabric and human body while significantly reducing computational overhead.

[0049] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0050] like Figure 1 As shown, the present application provides a collision detection method in clothing simulation, comprising:

[0051] Step 101: Acquire multiple action sequences provided by a subject with specified morphological parameters; each action sequence corresponds to an action; and each action sequence includes posture parameters of multiple frames.

[0052] In step 101, the action sequence is obtained through the DFaust subset of the AMASS dataset.

[0053] Step 102: For the posture parameters of each frame, input the posture parameters into the SMPL model to obtain the SMPL model under the posture parameters.

[0054] Step 103: Based on the SMPL model under each posture parameter, obtain the sampling space corresponding to each frame.

[0055] Step 104: For the sampling space corresponding to each frame, multiple different position points are sampled in the sampling space, half of which are obtained by spatially uniform sampling, and the other half are obtained by dense Poisson disk sampling on the human body surface.

[0056] Step 105: Construct a training data set using all the position points sampled in the sampling space corresponding to each frame, and whether each position point is inside or outside the human body's occupation field; where 1 indicates that the position point is inside the human body's occupation field, and 0 indicates that the position point is outside the human body's occupation field.

[0057] Step 106: Construct an MLP-based skin weight field prediction network and an implicit geometric representation deep learning network, including an MLP network for the skin weight field and an implicit geometric representation occupancy network for the human body under T-Pose, to obtain a collision detection model.

[0058] In step 106, the MLP network for the skinned weight field consists of a 4x128 MLP followed by a softmax module, using softplus as the activation function. The occupancy network for the implicit geometric representation of the human body in T-Pose consists of an 8x256 MLP followed by a sigmoid module, using softplus as the activation function. The 8x256 MLP has a skip connection at the fifth layer.

[0059] Step 107: The collision detection model is trained using the training data set to obtain a trained collision detection model; the MLP network of the skin weight field in the trained collision detection model is used to predict the skin weight of the input position point to be detected using a differentiable forward skinning network, and the T-Pose space corresponding point is obtained based on the skin weight combined with the Broyden method; the implicit geometric representation occupancy network of the human body under T-Pose in the trained collision detection model is used to determine whether the position point to be detected is inside or outside the human body's occupancy field based on the input T-Pose space corresponding point.

[0060] Step 108: Obtain the position of the vertex of the clothing to be detected.

[0061] Step 109: Use the trained collision detection model to determine whether the vertex of the clothing to be detected is inside or outside the occupied field of the human body. If the vertex of the clothing to be detected is outside the occupied field of the human body, then no collision occurs between the cloth and the human body. If the vertex of the clothing to be detected is inside the occupied field of the human body, then a collision occurs between the cloth and the human body. Move the collided vertex in the opposite direction of the gradient to resolve the collision.

[0062] In step 109 , the gradient is obtained through the trained collision detection model.

[0063] The technical solution of this application is described below with a specific embodiment:

[0064] The collision detection method in clothing simulation provided in this application is a clothing simulation collision processing method based on implicit human body probability field. It uses a neural network to predict the occupancy field of a digital human model with specific body parameters (specified morphological parameters) under corresponding posture parameters. The output value of the occupancy probability field and its gradient are used for collision detection and collision processing in physical simulation.

[0065] like Figure 2As shown, the clothing simulation collision processing method based on the implicit human body probability field of the present application includes the following steps:

[0066] 1. Obtain training motion sequences, partition the dataset, and process it into the occupancy fields required for training. Dataset Selection and Partitioning: The training data is the DFaust subset of the AMASS dataset, which contains 10 subjects of varying body shapes. Each subject provides 10 motion sequences. Model Training and Testing: One of the 10 motion sequences is randomly selected for validation, and the remaining 9 are used for training. The test data is the PosePrior subset of the AMASS dataset, which contains more challenging natural poses not covered by DFaust and is used to evaluate the model's generalization ability to new poses.

[0067] 2. Data Sampling and Processing: Occupancy field data for the SMPL model is generated using the action sequences and specified morphological parameters in the dataset. For each frame, 20,000 points are sampled, half using spatially uniform sampling and the other half using Poisson disk sampling on the mesh surface. Isotropic Gaussian noise (σ = 0.01) is then added to the sampled points to obtain points near the surface. Joint and bone rotation matrix information is then registered using the accompanying SMPL model. This section describes how to generate the aforementioned occupancy field data by sampling near the bounding box of the human body. The specified morphological parameters (body parameters) are SMPL terminology, typically 10 numbers. These morphological parameters are tied to the geometry of the human body being processed and are manually tuned or optimized in advance. This depends on the human body model used by the algorithm user for cloth simulation. The occupancy field data for the SMPL model consists of a series of coordinate values ​​(x, y, z, 0 or 1), plus a set of points that determine whether they are occupied or not. This set of points constitutes the occupancy field data used during training.

[0068] 3. Build a deep learning network, namely, an MLP-based skin weight field prediction network and an implicit geometry representation deep learning network. This includes an MLP network for the skin weight field and an MLP network for the static shape field of T-pose (i.e., an implicit geometry representation occupancy network for the human body under T-pose). The two networks are distributed across two modules, as described below:

[0069] The MLP network of the skin weight field does not improve the MLP network, but only uses the MLP network structure to fit and predict the skin weight field. The network structure diagram of the two neural networks based on MLP in this application is as follows Figure 3 and Figure 4 As shown, Figure 3 The detailed structure diagram of the MLP network of the skin weight field is given in the figure. The corresponding parameters and activation functions are in Figure 3 It points out that Figure 4The specific structure diagram of the static shape field MLP network of T-pose is given in the figure, and the corresponding parameters and activation functions are in Figure 4 It is pointed out in.

[0070] Model solution process:

[0071] (1) The input is the spatial coordinate x' to be queried and the geometric posture parameter θ to be queried. i .

[0072] During the simulation, the coordinates of the points to be queried are the spatial coordinates of the cloth vertices in each frame of the cloth simulation. These coordinates are then passed to the network to determine whether a collision has occurred with the human body. Pose parameters, which refer to the sequence of poses in the human animation simulation, are input to the algorithm and are provided by the user. In this case, they represent the rotation of the human skeleton expressed as an axis angle.

[0073] (2) The task of the first module (i.e., the MLP network of the skin weight field) is to find the corresponding point x in the T-Pose space before the query coordinate point x' is linearly skinned. * , in order to complete this task, the following equation is solved by Broyden iterative root finding method:

[0074]

[0075] where x * is the solution to the equation and the input to the next module, i.e., the corresponding point of the input point x' in the T-Pose space. Since iterative root search starts from different initial points, there will be multiple possible solutions in the end. All possible solutions are expressed as arrive n b Represents the number of skeletal joints in the human body, B i Represents the rotation matrix corresponding to the i-th bone joint, Represents the skin weight predicted by the network corresponding to the i-th bone joint, What is calculated is the linear skin corresponding to the current iterative solution.

[0076] The specific solution process is to first set the initial position x of the solution * By querying w The skin weight network with parameters is obtained That is, the skin weight value required for forward skinning (this is the role of the first network, input a point, output its corresponding skin weight, used to calculate linear skinning), by calculating Get the difference between the solution of the current iteration and the target. When the difference is less than a certain threshold ∈, it indicates that the iteration has converged. It is set to 10 -5 At this time arrive is the output of the first module and is used as the input of the next network, which means the corresponding points of multiple possible T-pose spaces of the query coordinate point.

[0077] (3) In the second module (i.e., the implicit geometric representation of the human body under T-Pose occupies the network, using Indicates), that is, the implicit occupancy field under T-Pose, subscript σ f Represents the parameters of the network module, by calculating each corresponding point arrive Probability value in the occupied field arrive where θ i The pose parameters of the SMPL parameterized human body model, i.e., the axis-angle representation of skeletal rotation, are fed as supplementary input to the second module. The maximum value, max, is taken as the output of the entire system. This represents the probability that the input point coordinate x' occupies the human body geometry in the current pose. When the output value is greater than or equal to 0, it indicates that the input query point is within or on the human body's occupancy field, indicating a collision. When the output value is less than 0, it indicates that the input query point is outside the human body's occupancy field, indicating no collision.

[0078] The second module does not involve iteration and does not need to return to step (1). After the second module is completed, the collision judgment result is obtained. The first module uses an iterative method to solve the transformation of the point coordinates of the two spaces. When the solution converges, the Broyden iterative root-finding algorithm used in module 1 (i.e., the first module) will stop iterating.

[0079] Training loss definition: The model is trained by minimizing the binary cross entropy loss between the predicted occupancy and the true occupancy of the corresponding training data.

[0080] for Figure 5 Explanation: For the training of the first network (i.e. the first module), the calculation of the gradient needs to be manually specified. During the training process, it is necessary to determine the overall loss L with respect to the skin weight network parameter σ={σ f ,σ w}, σ f represents the learnable parameters of the network in module 2 (i.e., the second module), σ w Represents the learnable parameters of the network in module 1. The gradient is expressed as:

[0081]

[0082] Where o represents the predicted occupancy output by the entire network.

[0083] The gradient described by the above formula can be easily obtained by backpropagating the gradient in the corresponding computational graph (computational graph refers to a data structure in training frameworks such as PyTorch and TensorFlow).

[0084] The parameters σ that need to be learned for the skin weight field w , the gradient is expressed as:

[0085]

[0086] where x * It is given by the equation The root defined, ∈ is the convergence threshold, set to Represents the output value x of the second module to the first module * Find the partial derivative, the last term It can be obtained by the following implicit differentiation. To simplify the writing, Recorded as Represents x * With B as the bone rotation, σ w The skin weights obtained by the network parameter Perform linear skinning (LBS).

[0087]

[0088]

[0089] The derivative relationship provided by the above formula ensures that the gradient of the overall loss can be calculated effectively, thereby achieving effective training and optimization of the neural network. Figure 5 shown.

[0090] The prediction of the skin weight in the first module uses a 4*128 MLP followed by a softmax module for training. During inference, it is sampled into a discrete grid representation, and the weights of the pre-calculated linear mixed skin are also stored in the grid.

[0091] Because the main speed bottleneck of inference is computing the equations at each iteration of Broyden's method. Computing the equations is time-consuming because it involves querying the skinning weights, which are parameterized by a multi-layer perceptron (MLP) during training, and then computing the linear blend skinning (LBS).

[0092] Because the skin weight field does not contain high-frequency details, the skin weight field is reparameterized using a low-resolution voxel grid. Skin weights are defined for each grid point. Trilinear interpolation is then used to obtain skin weights for any non-grid-aligned point in space. A resolution of 64×64×16 is sufficient to describe the skin weights. Due to the "flatness" of the human body along the z-axis in typical spatial space, a lower resolution is used along the z-axis. Using a voxel representation rather than a neural field representation improves computational efficiency; a single trilinear interpolation requires only eight read operations and eliminates the need for an MLP.

[0093] The second module is the implicit geometric representation of the human body under T-Pose, which is an 8*256 MLP. Finally, a sigmoid function is connected and softplus is used as the activation function. The 8*256 MLP will make a skip connection at the 5th layer, using high-level semantic information and low-level details to alleviate the gradient disappearance problem caused by increasing the depth of the deep neural network. Figure 4 shown.

[0094] 4. In physical simulation, it is necessary to determine whether an object such as clothing collides with the human body. By inputting the vertex coordinates of the object in the current frame and the posture information of the model in the current frame into the network inference, the occupancy probability information and gradient information can be obtained. Figure 6 As can be seen in the figure, the entire system can be divided into two modules. The first module is used to find the corresponding point of the current query point in the T-Pose space. The second module is an implicit expression network of the human body geometry under T-Pose. By using the corresponding point to query the network, it can be determined whether the coordinates of the corresponding point are inside the human body geometry, thereby determining whether the current query point has collided.

[0095] Occupancy probability information is used to determine whether a collision has occurred. A value greater than or equal to 0 indicates a collision has occurred, while a value less than 0 indicates no collision has occurred. Collisions can be resolved by using the gradient information obtained during the query to move the collided vertices in the opposite direction of the gradient, thereby resolving the collision.

[0096] Differentiable forward skinning in the first module:

[0097] To predict the occupancy probability o of an observed 3D point x' in the deformed space x' , we must first determine the canonical corresponding point x of the deformed query x' * , in order to evaluate the occupancy rate using a canonical occupancy network.

[0098] The core of the forward skinning method is to find the canonical corresponding point x for any query point x' * Because (i) they are related by the equation is implicitly defined and has no analytical inverse form, and (ii) multiple canonical points may correspond to the same deformation point because the spaces may overlap after deformation. To address this problem, an iterative optimization method is used, which is a method that can retrieve all potential canonical corresponding points for any deformation point x' from the implicitly defined relationship These corresponding points are then composited using standard operations of implicit shape composition.

[0099] Corresponding point search: Forward skinning implicitly defines the canonical corresponding point x of x' * is the root of the following equation:

[0100]

[0101] This equation cannot be solved in closed form. The solution can be obtained by iterative solution using standard Newton or quasi-Newton numerical solvers, as follows:

[0102]

[0103] Among them, J k yes The Jacobian matrix of the kth iteration, x k+1 and x k are the iterative solutions for the k+1th and kth rounds respectively. To avoid computing the Jacobian at each iteration, Broyden's method is applied using a low-rank approximation of J.

[0104] Use Broyden's iteration to find roots:

[0105] In parallel with the query point x', the input is a precomputed transformed voxel grid {T v The core first uses the bone transformation B and the query point x' to calculate multiple initialization states.

[0106] In each root-finding iteration, the core trilinear interpolation transforms the point x' and applies the Broyden update. After obtaining the iterative structure for each initial value, another important speed optimization is related to the handling of multiple correspondences found by the root-finding algorithm. The set of valid correspondences contains duplicates because different initial states can converge to the same solution. To avoid unnecessary evaluation of these duplicate canonical neural fields, duplicate solutions are detected by their relative distance in the canonical space and discarded.

[0107] 5. The corresponding points in the optimized T-Pose space arrive Passed into the second module, calculated through the second network (i.e. the second module) arrive The corresponding occupancy probability and gradient information are obtained, and the maximum probability value is taken as the final output of the entire system, indicating the probability of whether the query point collides with the human body in the current posture.

[0108] 6. In practical applications, the query points are the coordinates of clothing vertices during simulation. The method provided in this application can be used to determine whether the fabric and the human body have collided. For example, clothing simulation algorithms often require detecting and resolving collisions between the fabric and the human body, often referred to as penetration. The method provided in this application can serve as a replacement module for the collision detection module in existing common simulation algorithms.

[0109] For example, by modeling the entire simulation process as an optimization problem, the method of the present application can be used to provide the energy term generated when a collision occurs, as well as the gradient used in the corresponding optimization. This optimization problem can be solved by conventional optimization methods such as gradient descent, and the position of the clothing point in the next frame can be obtained, completing the solution of the cloth simulation, thereby solving the collision detection and collision resolution problems between clothing and the human body in cloth simulation in the animation and game industries. The system is used as a discriminator to determine whether a cloth vertex collides with the human body in the current frame. The system inputs the probability of whether a query point collides and the gradient information of the corresponding point. The positive and negative values ​​and size of the probability can determine the occurrence of a collision, and the gradient information can be used to resolve the collision.

[0110] The present application relates to the dynamic reconstruction and animation of non-rigid objects, such as human models, for use in three-dimensional animation and virtual reality applications. With the development of technologies such as virtual reality, augmented reality, and the metaverse, the demand for efficient and accurate reconstruction and animation of non-rigid, articulated three-dimensional objects (such as the human body) has increased dramatically. These application scenarios require the ability to automatically infer the three-dimensional form and deformation of the human body in various postures without pre-set postures. Traditional methods often rely on manual annotation or specific data input, which limits their flexibility and wide application. The limitations of existing technologies mainly lie in the low computational efficiency and poor adaptability to dynamic posture changes when dealing with non-rigid objects, especially in the modeling and animation of the dynamic form of the human body. For example, although neural field-based technologies can handle complex geometric transformations, they still face high computational costs in real-time applications. In addition, these methods often perform poorly when dealing with unseen postures because they usually require seeing all possible posture variations during the training phase. The algorithm proposed in this application significantly improves computational efficiency while maintaining accuracy. By introducing a voxel-based skin weight field representation in the neural field and pre-calculating the linear blend skinning (LBS) function, the algorithm improves the deformation modeling method from the posture space to the canonical space, and realizes efficient three-dimensional human body dynamic simulation. In addition, the algorithm supports unsupervised learning and does not need to rely on pre-labeled skin weights and other data, which significantly improves the model's generalization ability for new postures, and supports the output of the corresponding query point gradient, which can better adapt to the collision detection algorithm in clothing simulation. The method of the present application can mainly be used for collision detection, and since the network can also obtain gradient information, it can be adapted to clothing simulation algorithms that need to use gradient information as collision resolution.

[0111] In summary, this application solves the problems of low efficiency and poor adaptability of previous neural field technologies in dynamic human body model reconstruction and animation generation by combining voxel grids and pre-calculation technology, and innovatively uses it in the collision processing module in clothing simulation to solve the problem of collision detection and resolution between fabric and human body. It provides a new solution for real-time three-dimensional animation generation in the fields of virtual reality, augmented reality and metaverse. Among them, the low efficiency and poor adaptability of previous neural field technologies in dynamic human body model reconstruction and animation generation refer to: previous neural field technologies are usually used in the field of three-dimensional reconstruction, with the purpose of obtaining a geometric representation of an object (especially static objects), and are not specially designed for the simulation field. Some existing neural field technologies that can obtain dynamics are inefficient in the task of realizing the geometric representation of dynamic objects and have poor adaptability to new action postures, that is, they cannot be generalized to new postures and actions. The innovative use of it in the collision processing module in clothing simulation means that: neural field technology is usually used for reconstruction tasks, which are usually for static objects. The existing methods are not effective in directly generating dynamic neural fields. The algorithm proposed in this application can solve the problem of poor adaptability of neural fields to dynamic objects, and innovatively apply this dynamic neural field technology to the field of clothing simulation, using dynamic neural fields to solve the problem of collision detection and resolution between fabrics and the human body. Previous fabric simulation systems did not use this method.

[0112] This application aims to solve the key technical difficulties in the current three-dimensional animation generation, namely how to use neural networks to learn and implement (construct) an implicit human body representation network driven by human body posture parameters, and then perform efficient collision processing in physical simulation. To achieve the above purpose, the present application provides a forward skinning method based on implicit neural networks for animating non-rigid neural implicit shapes, such as human body models. This method combines the advantages of linear blend skinning (LBS) of traditional polygonal meshes and neural implicit surfaces, and can learn forward deformation fields without predefined skinning weights. The present application uses iterative root finding technology to find the corresponding points of all canonical spaces of any deformation point, and implements analytical gradients through implicit differentiation, supporting end-to-end training from 3D mesh and bone transformations. The present application provides the following solutions:

[0113] A clothing simulation collision processing method based on an implicit human body probability field, the method comprising:

[0114] It is very difficult to use skeletal information as input encoding to achieve implicit representation of the dynamic human body. The reason is that skeletal information represented by axis-angle is a very discrete and non-compact expression. In 3D, two poses that look very similar to the human eye may be very far apart in the space where the encoding is located. In addition, controlling the output of the network only by xyz coordinates will also increase the burden on the network. These two factors will make it difficult for the network training to converge and difficult to have good generalization. Among them, the network refers to the set of algorithms in this application. This set of algorithms can be understood as a function that inputs some xyz and posture information and outputs whether a collision occurs and the corresponding gradient. The dynamic neural field is realized through a neural network, and the network is used here to refer to the dynamic neural field.

[0115] The key to solving these problems lies in modeling the deformation relationship between 3D positions in the static pose (a canonical space) and the deforming space, thereby reducing the network's burden and convergence difficulty. This method uses iterative root search to find an accurate correspondence between the canonical space and the pose space. This decouples the pose parameters from the network's representation, requiring the network to accurately represent the geometry in the canonical space. The 3D position refers to the 3D spatial position of the query point. During training, these 3D point positions are sampled within the bounding box of the human body geometry. In cloth simulation, the 3D position of the cloth must be simulated. The 3D coordinates of the query points used as input to the algorithm are the vertices of the cloth. Collision detection involves determining whether these vertices collide with the dynamic neural field of the human body during simulation, a phenomenon known as "passing through the model." Correspondence refers to the relationship between the same query point in different coordinate systems. For a query point in pose space, its position in the canonical space must be found. These two positions are converted from the canonical space to the pose space via a linear skinning transformation.

[0116] What needs to be done here is the inverse process, which requires iterative root-finding calculations to obtain the corresponding relationship.

[0117] We first define the representation of the canonical shape and the forward skinning weights. We then introduce the forward mapping and derive the gradients for learning the canonical shape representation and the skinning weights from start to finish.

[0118] Representation: Represent an articulated object in terms of shape and skinning weights in a canonical space. In a manner similar to classical methods such as SMPL, the problem (how to learn a dynamic neural field through a network) is divided into pose-independent linear blend skinning (LBS) and pose-dependent nonlinear deformations. LBS captures many important aspects of shape variation, so pose-dependent models only need to learn a single correction term. This enables training with limited data and helps generalize to unseen poses. Shape: This application uses a neural network to predict the occupancy probability of any input 3D point in a canonical space. In order to model pose-dependent local deformations, such as wrinkles or soft tissue, the object pose is taken as an additional input: Among them, σ f is the network parameter, n p It is the posture condition The dimensions of , specified in joint angles, Denotes the real space, and the superscripts refer to the spatial dimensions. The canonical shape is implicitly defined as the 0.5 level set of the neural function S:

[0119] Neural Hybrid Skinning: Uses a linear blend skinning (LBS) model to simulate non-rigid deformations caused by skeletal changes. A second neural network is used to represent an LBS weight field in the canonical space, denoted by w:

[0120]

[0121] Among them, σ w is the network parameter, n b Indicates the number of bones. Following the traditional LBS, the softmax activation function is used to ensure the weight of each point x Satisfy w i ≥0 and ∑ i w i =1,w i is the weight corresponding to the bone index i, Independent of the pose θ i .

[0122] Given a 3D point x, the LBS weight w corresponds to a specific posture θ i Bone transformation The inflection point x is determined by the following convex combination:

[0123]

[0124] Handling multiple correspondences: Finding multiple roots by initializing the optimization process with different initial positions and exploiting the local convergence of Broyden's method for iterative root finding Initial state Obtained by:

[0125]

[0126] The final corresponding point set X * Defined as:

[0127]

[0128] where ∈ is the convergence threshold set in the experiment. This allows retrieving all canonical corresponding points for any deformed point x′ under topological changes. If any canonical corresponding point is occupied, the deformed point x′ is also considered occupied. Therefore, the maximum occupancy probability of all canonical corresponding points gives the final occupancy prediction:

[0129]

[0130] This joint operation is often used to synthesize independent shapes. In practice, softmax is used instead of max to allow the gradient to propagate back to all canonical corresponding points.

[0131]

[0132] The gradient of the output at each point This can be achieved by numerically estimating the gradient.

[0133] A better approach is to use automatic differentiation packages. This approach allows the required gradients to be calculated directly through the neural network structure without the need for complex numerical estimation, thereby improving overall computational efficiency and accuracy. The obtained gradient values ​​allow the model output to have a wider range of uses. Figure 7 shown.

[0134] The form of each layer of MLP is:

[0135] y l+1 =σ(Wy l +b);

[0136] in, is a nonlinear differentiable activation function (Softplus is used in this application), W,b are the learnable parameters of this layer, y l is the input of layer l, y l+1 is the output of the lth layer and the input of the l+1th layer.

[0137] According to the chain rule, the gradient satisfies the following relationship:

[0138]

[0139] Among them, diag(z) means inputting the vector The operations are arranged on the diagonal of a k×k square matrix, σ' is the derivative of σ, and are the gradients of the lth and l+1th layers of the MLP with respect to the input x, respectively.

[0140] equation show, It can be constructed as a neural network combined with f(x;θ), where f(x;θ) represents the function mapping represented by the neural network with θ as the learnable parameter and x as the input. Represents the gradient of the input x with respect to the function.

[0141] For t=0, that is, the gradient of the first layer, there is usually no gradient of the previous layer when the network is initialized. So, we need Set an initial value. In the first layer, y 0 Typically this is the input x itself (without any complex transformations done by the network’s input layer), so: here Represents the identity matrix, whose dimensions are the same as the input x.

[0142] Thus, the initial value of the gradient simply means that the derivative of each input element with respect to itself is 1, and the derivative with respect to other input elements is 0. Therefore, for the first layer:

[0143]

[0144] in, represents the gradient of the first layer.

[0145] The following figure provides an overview of the clothing simulation collision processing algorithm based on implicit human probability field. Figure 8 As shown, existing cloth simulation algorithms do not use neural networks to represent collision bounding boxes. This application uses neural networks to learn an occupancy field that can be transformed according to human posture, which has the following advantages:

[0146] 1. Usually the two-stage collision detection can be compressed into a single neural network query.

[0147] 2. The collision field constructed by spatial hashing needs to be recalculated and reconstructed to adapt to the new posture. Using a trained neural network does not require this step. It only needs to change the posture parameters of the neural network input. Because the trained neural network can generalize to new postures, it can save computational overhead.

[0148] 3. The neural network represents a continuous function, so the accuracy of collision query is theoretically infinite, and the geometric resolution of the human body obtained by fitting is much higher than the collision field obtained by discrete space hashing.

[0149] Based on the same inventive concept, embodiments of the present application also provide a collision detection system for clothing simulation, which is used to implement the aforementioned collision detection method for clothing simulation. The solution provided by this system is similar to the solution described in the aforementioned method. Therefore, the specific limitations of one or more embodiments of the collision detection system for clothing simulation provided below can be found in the aforementioned limitations of the collision detection method for clothing simulation, and will not be further elaborated here.

[0150] In an exemplary embodiment, a collision detection system in clothing simulation (i.e., a clothing simulation collision processing system based on an implicit human body probability field) is provided, comprising:

[0151] The action sequence acquisition module is used to acquire multiple action sequences provided by subjects with specified morphological parameters; each action sequence corresponds to an action; and each action sequence includes posture parameters of multiple frames.

[0152] The SMPL model acquisition module under each posture parameter is used to input the posture parameters into the SMPL model for the posture parameters of each frame to obtain the SMPL model under the posture parameters.

[0153] The sampling space acquisition module corresponding to each frame is used to obtain the sampling space corresponding to each frame based on the SMPL model under each posture parameter.

[0154] The different position point sampling module is used to sample multiple different position points in the sampling space corresponding to each frame, half of which are obtained by spatial uniform sampling, and the other half are obtained by dense Poisson disk sampling on the human body surface.

[0155] The training data set construction module is used to construct a training data set using all the position points sampled in the sampling space corresponding to each frame, as well as whether each position point is inside or outside the human body's occupation field; wherein 1 indicates that the position point is inside the human body's occupation field, and 0 indicates that the position point is outside the human body's occupation field.

[0156] The collision detection model construction module is used to build an MLP-based skin weight field prediction network and an implicit geometric representation deep learning network, including an MLP network for the skin weight field and an implicit geometric representation occupancy network for the human body under T-Pose, to obtain a collision detection model.

[0157] The collision detection model training module is used to train the collision detection model using the training data set to obtain a trained collision detection model; the MLP network of the skin weight field in the trained collision detection model is used to predict the skin weight of the input position point to be detected using a differentiable forward skinning network, and the T-Pose space corresponding point is obtained based on the skin weight combined with the Broyden method. The implicit geometric representation occupancy network of the human body under T-Pose in the trained collision detection model is used to determine whether the position point to be detected is inside or outside the human body's occupancy field based on the input T-Pose space corresponding point.

[0158] The clothing vertex position acquisition module is used to obtain the position of the vertices of the clothing to be detected.

[0159] The collision detection module is used to use the trained collision detection model to determine whether the vertex of the clothing to be detected is inside or outside the human body's occupation field. If the vertex of the clothing to be detected is outside the human body's occupation field, then there is no collision between the cloth and the human body. If the vertex of the clothing to be detected is inside the human body's occupation field, then there is a collision between the cloth and the human body. The collided vertex is moved in the opposite direction of the gradient to resolve the collision.

[0160] Among them, the action sequence is obtained through the DFaust subset of the AMASS dataset.

[0161] The MLP network of the skinned weight field consists of a 4*128 MLP and a Softmax module. The 4*128 MLP is followed by a softmax module, and softplus is used as the activation function.

[0162] The implicit geometric representation of the human body under T-Pose occupies a network consisting of an 8*256 MLP and a Sigmoid module. The 8*256 MLP is followed by the Sigmoid module and uses softplus as the activation function. The 8*256 MLP will make a skip connection at the 5th layer.

[0163] The gradient is obtained through the trained collision detection model.

[0164] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 9As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store collision detection data in clothing simulation. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a collision detection method in clothing simulation is implemented.

[0165] Those skilled in the art will understand that Figure 9 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present application and does not constitute a limitation on the computer device to which the solution of the present application is applied. A specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps of the above-mentioned method embodiments when executing the computer program.

[0166] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0167] In an exemplary embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0168] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0169] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0170] The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processors involved in the various embodiments provided herein may include, but are not limited to, general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic units, data processing logic units based on quantum computing, and the like.

[0171] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0172] This document uses specific examples to illustrate the principles and implementation methods of this application. The description of the above examples is only intended to help understand the method and core concept of this application. At the same time, for those skilled in the art, based on the concept of this application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A collision detection method in clothing simulation, characterized in that: The collision detection method in the clothing simulation includes: Acquire multiple action sequences provided by a subject with specified morphological parameters; each of the action sequences corresponds to an action; each of the action sequences includes posture parameters of multiple frames; For each frame's posture parameter, input the posture parameter into the SMPL model to obtain the SMPL model under the posture parameter; Based on the SMPL model under each posture parameter, obtain the sampling space corresponding to each frame; For each frame, a plurality of different position points are sampled in the sampling space, wherein half of the position points are obtained by spatially uniform sampling and the other half of the position points are obtained by dense Poisson disk sampling on the human body surface; A training dataset is constructed using all the position points sampled in the sampling space corresponding to each frame, and whether each position point is inside or outside the human body's occupancy field; 1 indicates that the position point is inside the human body's occupancy field, and 0 indicates that the position point is outside the human body's occupancy field; Construct an MLP-based skin weight field prediction network and an implicit geometry representation deep learning network, including an MLP network for the skin weight field and an implicit geometry representation occupancy network for the human body under T-Pose, to obtain a collision detection model. The collision detection model is trained using the training data set to obtain a trained collision detection model; an MLP network of a skin weight field in the trained collision detection model is used to predict skin weights for an input position point to be detected using a differentiable forward skinning network, and a T-Pose space corresponding point is obtained based on the skin weights in combination with a Broyden method; an occupancy network of an implicit geometric representation of the human body under T-Pose in the trained collision detection model is used to determine whether the position point to be detected is inside or outside the human body's occupancy field based on the input T-Pose space corresponding point; Get the position of the vertex of the clothing to be detected; The trained collision detection model is used to determine whether the vertex of the clothing to be detected is inside or outside the occupation field of the human body. If the vertex of the clothing to be detected is outside the occupation field of the human body, no collision occurs between the cloth and the human body. If the vertex of the clothing to be detected is inside the occupation field of the human body, a collision occurs between the cloth and the human body. The collided vertex is moved in the opposite direction of the gradient to resolve the collision.

2. The collision detection method in clothing simulation according to claim 1, characterized in that: The action sequence is obtained through the DFaust subset of the AMASS dataset.

3. The collision detection method in clothing simulation according to claim 1, characterized in that: The MLP network of the skin weight field includes a 4*128 MLP and a Softmax module. The 4*128 MLP is followed by a softmax module, and softplus is used as the activation function.

4. The collision detection method in clothing simulation according to claim 1, characterized in that: The implicit geometric representation of the human body under T-Pose occupies a network consisting of an 8*256 MLP and a Sigmoid module. The 8*256 MLP is followed by the Sigmoid module, using softplus as the activation function. The 8*256 MLP will make a skip connection at the 5th layer.

5. The collision detection method in clothing simulation according to claim 1, characterized in that: The gradient is obtained through the trained collision detection model.

6. A collision detection system in clothing simulation, characterized in that: The collision detection system in the clothing simulation includes: An action sequence acquisition module is used to acquire multiple action sequences provided by a subject with specified morphological parameters; each action sequence corresponds to an action; and each action sequence includes posture parameters of multiple frames; An SMPL model acquisition module under each posture parameter is used to input the posture parameter into the SMPL model for each frame posture parameter to obtain the SMPL model under the posture parameter; The sampling space acquisition module corresponding to each frame is used to obtain the sampling space corresponding to each frame based on the SMPL model under each posture parameter; A different position point sampling module is used to sample multiple different position points in the sampling space corresponding to each frame, where half of the position points are obtained by spatially uniform sampling and the other half are obtained by dense Poisson disk sampling on the human body surface; A training data set construction module is used to construct a training data set using all the position points sampled in the sampling space corresponding to each frame, and whether each position point is inside or outside the human body's occupation field; where 1 indicates that the position point is inside the human body's occupation field, and 0 indicates that the position point is outside the human body's occupation field; The collision detection model construction module is used to build an MLP-based skin weight field prediction network and an implicit geometric representation deep learning network, including an MLP network for the skin weight field and an implicit geometric representation occupancy network for the human body under T-Pose, to obtain a collision detection model. A collision detection model training module is configured to train the collision detection model using the training data set to obtain a trained collision detection model; an MLP network of the skin weight field in the trained collision detection model is configured to predict skin weights for input positions to be detected using a differentiable forward skinning network, and to obtain corresponding points in T-Pose space based on the skin weights in combination with the Broyden method; an occupancy network of the implicit geometric representation of the human body under T-Pose in the trained collision detection model is configured to determine whether the position to be detected is inside or outside the human body's occupancy field based on the input corresponding points in T-Pose space; A clothing vertex position acquisition module is used to obtain the position of the vertex of the clothing to be detected; The collision detection module is used to use the trained collision detection model to determine whether the vertex of the clothing to be detected is inside or outside the occupied field of the human body. If the vertex of the clothing to be detected is outside the occupied field of the human body, then no collision occurs between the cloth and the human body. If the vertex of the clothing to be detected is inside the occupied field of the human body, then a collision occurs between the cloth and the human body. The collided vertex is moved in the opposite direction of the gradient to resolve the collision.

7. The collision detection system in clothing simulation according to claim 6, characterized in that: The gradient is obtained through the trained collision detection model.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the collision detection method in clothing simulation according to any one of claims 1 to 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the collision detection method in clothing simulation according to any one of claims 1 to 5 is implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the collision detection method in clothing simulation according to any one of claims 1 to 5 is implemented.

Images