Active Learning of Neural Collision Handler for Complex 3D Mesh Deformations
This addresses collision handling for complex 3D meshes in applications like human poses and cloth simulations, representing an incremental improvement over existing learning methods.
The paper tackles collision detection and handling in 3D deforming meshes by developing a learning algorithm that uses a bilevel deep autoencoder with attention and numerical optimization, achieving 93.8-98.1% accuracy compared to ground truth and outperforming prior methods with lower false negative and higher success rates.
We present a robust learning algorithm to detect and handle collisions in 3D deforming meshes. Our collision detector is represented as a bilevel deep autoencoder with an attention mechanism that identifies colliding mesh sub-parts. We use a numerical optimization algorithm to resolve penetrations guided by the network. Our learned collision handler can resolve collisions for unseen, high-dimensional meshes with thousands of vertices. To obtain stable network performance in such large and unseen spaces, we progressively insert new collision data based on the errors in network inferences. We automatically label these data using an analytical collision detector and progressively fine-tune our detection networks. We evaluate our method for collision handling of complex, 3D meshes coming from several datasets with different shapes and topologies, including datasets corresponding to dressed and undressed human poses, cloth simulations, and human hand poses acquired using multiview capture systems. Our approach outperforms supervised learning methods and achieves $93.8-98.1\%$ accuracy compared to the groundtruth by analytic methods. Compared to prior learning methods, our approach results in a $5.16\%-25.50\%$ lower false negative rate in terms of collision checking and a $9.65\%-58.91\%$ higher success rate in collision handling.