The complexity of editing realistic motion in 3D character animation poses significant challenges for animators, requiring both artistic and technical expertise. Current motion editing tools are often unintuitive and restrict immersive interaction with the animation process. This proposal introduces an innovative mixed-reality (MR) interface, leveraging Virtual Reality (VR) to enhance motion editing by allowing animators to manipulate character joints through intuitive hand gestures in an immersive environment. The system will provide real-time feedback on joint adjustments, showing their impact on the overall animation, including neighboring temporal points. By incorporating advanced graphical neural networks (GNNs) based on the latest attention-based designs, our approach addresses complex motion constraints and enables realistic, fluid motion editing. Additionally, the interface will simplify editing with hard constraints by letting animators directly adjust key poses along a virtual timeline, offering increased creative freedom and precision. This project aims to redefine animation authoring tools, making them more accessible for beginners and more efficient for professionals, ultimately advancing the future of animation technology.

My interest in my research in computer graphics and HCI stems from my passion for animation and storytelling. Traditionally, animation tools are often complex and hard to pick up. Therefore, I want to democratize animation and the art of storytelling by designing intuitive, streamlined interfaces that allow anyone to bring their creative visions to life. My research tackles this massive challenge from the motion editing perspective, leveraging the power of mixed reality to make animation more approachable and empower both beginners and professionals to tell their stories with greater ease and precision.