Table of Contents
Fetching ...

AnimAgents: Coordinating Multi-Stage Animation Pre-Production with Human-Multi-Agent Collaboration

Wen-Fan Wang, Chien-Ting Lu, Jin Ping Ng, Yi-Ting Chiu, Ting-Ying Lee, Miaosen Wang, Bing-Yu Chen, Xiang 'Anthony' Chen

TL;DR

AnimAgents tackles the fragmentation of animation pre-production by introducing a stage-aware human–multi-agent system that orchestrates four specialized agents (Ideation, Scripting, Design, Art) under a central Core Agent. It employs stage-specific boards and lineage-aware blocks to preserve continuity and enable end-to-end collaboration from concept to storyboard, supported by memory augmentation and a robust image-generation pipeline. In a summative within-subject study (n=16) against a strong single-agent baseline, AnimAgents significantly improved coordination, continuity, information management, traceability, and user satisfaction (p < .01), with field deployments (n=4) confirming real-world value, especially in commercial contexts. The work demonstrates how structured, agent-based coordination can preserve creative agency while increasing efficiency, suggesting broader applicability to other complex, multi-stage creative domains and highlighting design considerations for nonlinear workflows, long-horizon memory, and ethical adoption.

Abstract

Animation pre-production lays the foundation of an animated film by transforming initial concepts into a coherent blueprint across interdependent stages such as ideation, scripting, design, and storyboarding. While generative AI tools are increasingly adopted in this process, they remain isolated, requiring creators to juggle multiple systems without integrated workflow support. Our formative study with 12 professional creative directors and independent animators revealed key challenges in their current practice: Creators must manually coordinate fragmented outputs, manage large volumes of information, and struggle to maintain continuity and creative control between stages. Based on the insights, we present AnimAgents, a human-multi-agent collaborative system that coordinates complex, multi-stage workflows through a core agent and specialized agents, supported by dedicated boards for the four major stages of pre-production. AnimAgents enables stage-aware orchestration, stage-specific output management, and element-level refinement, providing an end-to-end workflow tailored to professional practice. In a within-subjects summative study with 16 professional creators, AnimAgents significantly outperformed a strong single-agent baseline that equipped with advanced parallel image generation in coordination, consistency, information management, and overall satisfaction (p < .01). A field deployment with 4 creators further demonstrated AnimAgents' effectiveness in real-world projects.

AnimAgents: Coordinating Multi-Stage Animation Pre-Production with Human-Multi-Agent Collaboration

TL;DR

AnimAgents tackles the fragmentation of animation pre-production by introducing a stage-aware human–multi-agent system that orchestrates four specialized agents (Ideation, Scripting, Design, Art) under a central Core Agent. It employs stage-specific boards and lineage-aware blocks to preserve continuity and enable end-to-end collaboration from concept to storyboard, supported by memory augmentation and a robust image-generation pipeline. In a summative within-subject study (n=16) against a strong single-agent baseline, AnimAgents significantly improved coordination, continuity, information management, traceability, and user satisfaction (p < .01), with field deployments (n=4) confirming real-world value, especially in commercial contexts. The work demonstrates how structured, agent-based coordination can preserve creative agency while increasing efficiency, suggesting broader applicability to other complex, multi-stage creative domains and highlighting design considerations for nonlinear workflows, long-horizon memory, and ethical adoption.

Abstract

Animation pre-production lays the foundation of an animated film by transforming initial concepts into a coherent blueprint across interdependent stages such as ideation, scripting, design, and storyboarding. While generative AI tools are increasingly adopted in this process, they remain isolated, requiring creators to juggle multiple systems without integrated workflow support. Our formative study with 12 professional creative directors and independent animators revealed key challenges in their current practice: Creators must manually coordinate fragmented outputs, manage large volumes of information, and struggle to maintain continuity and creative control between stages. Based on the insights, we present AnimAgents, a human-multi-agent collaborative system that coordinates complex, multi-stage workflows through a core agent and specialized agents, supported by dedicated boards for the four major stages of pre-production. AnimAgents enables stage-aware orchestration, stage-specific output management, and element-level refinement, providing an end-to-end workflow tailored to professional practice. In a within-subjects summative study with 16 professional creators, AnimAgents significantly outperformed a strong single-agent baseline that equipped with advanced parallel image generation in coordination, consistency, information management, and overall satisfaction (p < .01). A field deployment with 4 creators further demonstrated AnimAgents' effectiveness in real-world projects.

Paper Structure

This paper contains 69 sections, 11 figures, 4 tables.

Figures (11)

  • Figure 1: Typical traditional animation pre-production workflow, which progresses through five stages: Ideation, Story/Scripting, Design, Storyboard, and Animatic, before moving into production. The workflow is not strictly linear, as iterations and revisions may occur across stages.
  • Figure 2: AnimAgents' workflow from P22 in the summative study. The user began by initiating a conversation with the Core Agent, describing a story idea about “Dream Architect.” The workflow progressed through the stages of Planning, Ideation, Design, a second round of Ideation, Scripting, and Storyboard, ultimately resulting in a completed shot list and storyboard.
  • Figure 3: System architecture of AnimAgents. For stage management, the Core Agent determines the (C1) current stage based on user input, provides (A) stage-specific prompts, and (B) tracks progress based on agent outputs. Results are then displayed on (C2) stage-specific boards. For agent management, it (D) delegates tasks to Specialized Agents and (E) reviews their results before approval. Users interact with agents through text and (G) selected blocks and elements, and can (F) directly communicate with Specialized Agents.
  • Figure 4: Main interface of AnimAgents. (A) Chat panel, where users can chat with agents, interact through conversation, and upload images. (B) Board workspace on the right, consisting of four infinite canvases (boards)—Ideation, Design, Scripting, and Storyboard—with a progress bar at the top showing the current stage progress and project information panel. Users can toggle which boards are displayed (B1) and expand or collapse an individual board (B2).
  • Figure 5: Ideation board in AnimAgents. Each block contains multiple options (e.g., loglines) for user selection and can be collapsed, pinned, or expanded. When a block is modified or extended, a new branch is created, preserving previous content and enabling flexible, iterative exploration.
  • ...and 6 more figures