Researchers Introduce New Technique for Artists to Control Animations
Artists who work on animated movies and video games now have a new technique at their disposal, thanks to researchers from MIT. This method involves the generation of mathematical functions called barycentric coordinates, which determine the bending, stretching, and movement of 2D and 3D shapes. By using this tool, artists can choose functions that align with their creative vision, allowing them to have more control over the animations of characters like heroes and villains.
Traditionally, existing techniques for this problem have been inflexible, offering only one option for the barycentric coordinate functions of a particular animated character. This limitation means that artists would have to start from scratch every time they wanted to achieve a slightly different look. However, the new technique developed by the MIT researchers provides a more generalized approach that allows artists to design or choose among various smoothness energies for any shape. This flexibility empowers artists to preview and select the deformation that best suits their artistic preferences.
The researchers utilized a special type of neural network to model the barycentric coordinate functions. By incorporating the constraints directly into the network’s architecture, they ensured that the generated solutions are always valid. This approach frees artists from having to worry about the mathematical aspects of the problem and enables them to focus on designing interesting barycentric coordinates.
To bridge the gap between Möbius’ triangular barycentric coordinates and complex modern cages, the researchers employed virtual triangles that cover the shape and connect triplets of points on the cage’s exterior. The neural network then predicts how to combine these virtual triangles’ barycentric coordinates to create a more intricate yet smooth function. This method allows artists to experiment with different coordinate functions and refine them until they achieve the desired animation.
The researchers demonstrated that their technique produces more natural-looking animations compared to other approaches. For example, they showcased a cat’s tail that moves with smooth curves instead of folding rigidly near the cage’s vertices. In the future, the team aims to explore strategies for accelerating the neural network and develop an interactive interface that enables real-time iteration on animations.
Beyond its applications in the artistic realm, this technique holds potential in fields such as medical imaging, architecture, virtual reality, and computer vision. It could assist robots in understanding how objects move in the real world.
The research paper detailing this technique was authored by Ana Dodik, Oded Stein, Vincent Sitzmann, and Justin Solomon. Their work was recently presented at SIGGRAPH Asia and has received funding from various organizations including the U.S. Army Research Office, the U.S. Air Force Office of Scientific Research, and the National Science Foundation.Kindly read our copyright disclaimer here: https://cere-sync.com/dmca-copyrights-disclaimer/