ViP-Rig: Visual-Prompted Controllable Rigging
ViP-Rig is a framework that uses visual prompts to control rigging in 3D models. It works by injecting information from user-drawn skeletal and rigidity prompts into frozen pre-trained models. This allows for prompt-first rigging and result-guided editing, leading to more accurate skeleton recovery and skinning weights compared to traditional geometry-based methods.