Animated films such as Toy Story, The Lion King, and Spirited Away hold cherished childhood memories for people all around the world. While we fondly remember the adorable characters and catchy theme songs, we’re less likely to reflect on the complex computer animation processes that enabled these classics. Such processes, especially with 3D animation, are incredibly time-consuming, with intermediate steps involving modelling, defining skeletal joints and deformation parameters, posing, keyframe setting, and much more.
Monster Mash, a novel AI-powered 3D modelling and animation tool, aims to make these arduous 3D animation processes a whole lot easier. The proposed framework comes from researchers at Czech Technical University in Prague, Google Research, University of Washington and ETH Zurich, and enables users to create animated 3D models from a single view.
Building on the 1999 work Teddy: A Sketching Interface for 3D Freeform Design by Igarashi et al, Monster Mash is described as the first sketch-based tool that enables creating and animating a smooth, consistent 3D model from a single viewpoint within seconds.
While previous sketch-based tools tended to require a tedious part-based multi-view workflow with the explicit specification of an animation rig, the proposed method combines 3D inflation with a rigidity-preserving, layered deformation model, ARAP-L, to produce a smooth 3D mesh, which is immediately ready for animation. Previous methods also often required users to deal with unwanted inter-penetrations, an issue that is eliminated by the proposed single viewpoint approach.
While the tool could benefit professional animators, it is also very user-friendly, so even a novice animator can simply sketch out a character and it will be automatically converted into a soft, deformable 3D model. The interface enables users to animate their 3D characters by simply grabbing and moving them around in real-time.
Researchers evaluated their Monster Mash approach on a variety of 3D models, where it was able to create complex 3D animations from fewer strokes and control points and in a fraction of the time compared to current state of the art methods. The team noted limitations of their approach, such as limited control over the proportions in depth. Although inflation together with depth-ordering constraints usually leads to good depth proportions, the researchers believe that adding control from different viewpoints might be beneficial, and envision a future extension of the framework to enable merging input from multiple poses or viewpoints.
The paper Monster Mash: A Single-View Approach to Casual 3D Modeling and Animation is on the Czech Technical University DCGI website. The project page is here.
Analyst: Yuqing Li | Editor: Michael Sarazen; Yuan Yuan
This report offers a look at how China has leveraged artificial intelligence technologies in the battle against COVID-19. It is also available on Amazon Kindle. Along with this report, we also introduced a database covering additional 1428 artificial intelligence solutions from 12 pandemic scenarios.
Click here to find more reports from us.
We know you don’t want to miss any latest news or research breakthroughs. Subscribe to our popular newsletter Synced Global AI Weekly to get weekly AI updates.