LottieGPT: Tokenizing Vector Animation for Autoregressive Generation

[论文] LottieGPT: Tokenizing Vector Animation for Autoregressive Generation ## 论文概要 **研究领域**: cs.CV **作者**: Junhao Chen, Kejun Gao, Yuehan Cui, Mingze Sun, Mingjin Chen, Shaohui Wang, Xiaoxiao Long, Fei Ma, Qi Tian, Ruqi Huang, Hao Zhao **发布时间**: 2026-04-13 **arXiv**: [2604.11792](https://arxiv.org/abs/2604.11792) ## 中文摘要 尽管视频生成进展迅速，现有模型仍无法生成矢量动画——这是互联网上一种主导且极具表现力的多媒体形式。矢量动画具有与分辨率无关、紧凑、语义结构和可编辑参数化运动表示等优点，但当前生成模型仅在栅格空间中运行，因此无法合成它们。本文提出首个矢量动画token化和自回归生成框架。采用广泛部署的JSON动画标准Lottie，设计定制的Lottie Tokenizer将分层几何图元、变换和基于关键帧的运动编码为紧凑且语义对齐的token序列。构建LottieAnimation-660K数据集，微调Qwen-VL创建LottieGPT，能够从自然语言或视觉提示直接生成连贯、可编辑的矢量动画。 ## 原文摘要 Despite rapid progress in video generation, existing models are incapable of producing vector animation, a dominant and highly expressive form of multimedia on the Internet. Vector animations offer resolution-independence, compactness, semantic structure, and editable parametric motion representations, yet current generative models operate exclusively in raster space and thus cannot synthesize them. --- *自动采集于 2026-04-15* #论文 #arXiv #AI #小凯