动画
工作流程
计算机科学
结构化
多媒体
计算机动画
约束(计算机辅助设计)
人机交互
运动(物理)
计算机图形学(图像)
生产(经济)
弹丸
自动化
工程制图
帧(网络)
控制(管理)
出处
期刊:
[SPIE]
日期:2026-04-08
卷期号:: 52-52
摘要
Aquatic-animal popular-science content is scientifically rigorous but often difficult for young audiences to understand, while conventional animation workflows remain slow and costly to revise in museum contexts. This paper proposes a practical AIGC-enabled design pathway for aquatic-animal popular-science animation following a four-stage loop: content deconstruction, model training and constraint design, animation generation, and interactive optimization. Driven by aquarium education needs, ten representative aquatic animals were used to validate the workflow. Language-model-assisted text structuring converts fact sheets into concept maps and shot lists; diffusion-based generation with structural control supports consistent assets and ecological scenes; motion templates with rig constraints synthesize plausible swimming gaits; and telemetry with A/B testing refines pacing and interaction. Experimental results show a knowledge-transfer accuracy of 91.5%, a 62% reduction in production cycle time compared with a manual 2D baseline, and a 47% increase in youth audience interest. The pathway provides a replicable technical solution for science museums and aquariums and offers practical evidence for responsible AIGC adoption in natural-science communication.
科研通智能强力驱动
Strongly Powered by AbleSci AI