Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Spiritus: An AI-Assisted Tool for Creating 2D Characters and Animations

Qirui Sun, Yunyi Ni, Teli Yuan, Jingjing Zhang, Fan Yang, Zhihao Yao, Haipeng Mi

TL;DR

Spiritus addresses the high technical barrier to producing personalized 2D character animation by combining NLP-driven text-to-character generation with skeleton-based animation. The system pipelines text prompts through image generation, SAM-based segmentation, and a non-uniform dynamic mesh with unified rigging to support costume variation, then maps motion data via BVH to 2D space using PCA and diffusion-based motion synthesis. Key contributions include a hierarchical character generation framework, a web-based interface with scene orchestration, and cross-platform interoperability via Spine Runtime, enabling reusable animation assets. The work demonstrates reduced setup complexity, enhanced creative flexibility, and asset universality for rapid animated skit production.

Abstract

This research presents Spiritus, an AI-assisted creation tool designed to streamline 2D character animation creation while enhancing creative flexibility. By integrating natural language processing and diffusion models, users can efficiently transform natural language descriptions into personalized 2D characters and animations. The system employs automated segmentation, layered costume techniques, and dynamic mesh-skeleton binding solutions to support flexible adaptation of complex costumes and additional components. Spiritus further achieves real-time animation generation and efficient animation resource reuse between characters through the integration of BVH data and motion diffusion models. Experimental results demonstrate Spiritus's effectiveness in reducing technical barriers, enhancing creative freedom, and supporting resource universality. Future work will focus on optimizing user experience and further exploring the system's human-computer collaboration potential.

Spiritus: An AI-Assisted Tool for Creating 2D Characters and Animations

TL;DR

Spiritus addresses the high technical barrier to producing personalized 2D character animation by combining NLP-driven text-to-character generation with skeleton-based animation. The system pipelines text prompts through image generation, SAM-based segmentation, and a non-uniform dynamic mesh with unified rigging to support costume variation, then maps motion data via BVH to 2D space using PCA and diffusion-based motion synthesis. Key contributions include a hierarchical character generation framework, a web-based interface with scene orchestration, and cross-platform interoperability via Spine Runtime, enabling reusable animation assets. The work demonstrates reduced setup complexity, enhanced creative flexibility, and asset universality for rapid animated skit production.

Abstract

This research presents Spiritus, an AI-assisted creation tool designed to streamline 2D character animation creation while enhancing creative flexibility. By integrating natural language processing and diffusion models, users can efficiently transform natural language descriptions into personalized 2D characters and animations. The system employs automated segmentation, layered costume techniques, and dynamic mesh-skeleton binding solutions to support flexible adaptation of complex costumes and additional components. Spiritus further achieves real-time animation generation and efficient animation resource reuse between characters through the integration of BVH data and motion diffusion models. Experimental results demonstrate Spiritus's effectiveness in reducing technical barriers, enhancing creative freedom, and supporting resource universality. Future work will focus on optimizing user experience and further exploring the system's human-computer collaboration potential.

Paper Structure

This paper contains 10 sections, 5 figures.

Table of Contents

  1. INTRODUCTION
  2. USER INTERFACE AND INTERACTION
  3. Scene Orchestration Component
  4. Character Creation Interface
  5. Animation and Hand Component Interface
  6. IMPLEMENTATION
  7. Image Generation and Segmentation
  8. Skeleton and Dynamic Mesh Binding
  9. Animation Generation and Mapping
  10. CONCLUSION

Figures (5)

  • Figure 1: Spiritus is an AI-assisted creation tool that integrates natural language processing, image generation, and skeletal animation technologies. Spiritus aims to provide personalized creative support for creators focused on character and animation creation: (a) Character image generation, (b) Animation and Hand components, (c) Scene orchestration, (d) User works.
  • Figure 2: Scene orchestration component interface, editing tools include (a) loop/single play, keyframe, move, mirror, delete tools, (b) character animation selection, (c) scene switching, animation playback, recording, export tools; interaction methods include (d) dragging characters into the scene, (e) adding and deleting keyframe animation paths, (f) scene selection, (g) scene text input generation.
  • Figure 3: Character creation interface workflow, including: (a) system supplementary description, (b) editing texture details based on the outline, (c) texture drawing, (d) selecting character outline or editing texture outline for character generation, (e) generation complete.
  • Figure 4: Animation editing and hand component interface. (a) Action selection, (b) BVH file conversion to character animation, (c) Text input for AI-generated animation, (d) Hand prop text input, (e) Hand prop sketching, (f) Example interface for hand extension component.
  • Figure 5: Implementation Framework