# VISTA: Video-Injected Stylized Text-to-Animation

arXiv: https://arxiv.org/abs/2609.23817 · HTML: https://mlsanity.com/paper/2609.23817/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 10 of its 14 references that are in the map.
- First posted on arXiv 20 Sep 2026.

## Builds on (most cited first, 10 of 10)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Adding Conditional Control to Text-to-Image Diffusion Models](https://mlsanity.com/paper/2302.05543/) (2023)
- [ViViT: A Video Vision Transformer](https://mlsanity.com/paper/2103.15691/) (2021)
- [Human Motion Diffusion Model](https://mlsanity.com/paper/2209.14916/) (2022)
- [Executing your Commands via Motion Diffusion in Latent Space](https://mlsanity.com/paper/2212.04048/) (2022)
- [Rethinking Diffusion for Text-Driven Human Motion Generation: Redundant Representations, Evaluation, and Masked Autoregression](https://mlsanity.com/paper/2411.16575/) (2024)
- [Real-Time Style Modelling of Human Locomotion via Feature-Wise Transformations and Local Motion Phases](https://mlsanity.com/paper/2201.04439/) (2022)
- [SMooDi: Stylized Motion Diffusion Model](https://mlsanity.com/paper/2407.12783/) (2024)
- [Dance Like a Chicken: Low-Rank Stylization for Human Motion Diffusion](https://mlsanity.com/paper/2503.19557/) (2025)
- [StyleMotif: Multi-Modal Motion Stylization using Style-Content Cross Fusion](https://mlsanity.com/paper/2503.21775/) (2025)

## Abstract (arXiv)

We present VISTA, a two-stage framework for generating stylized 3D human motion by fusing structural content from text prompts with expressive style from reference videos, without requiring jointly paired (text, video, stylized motion) triplets. A Dual-channel Autoencoder first maps motion sequences and video clips into a shared latent manifold. A masked autoregressive diffusion backbone then operates within this manifold, injecting video-derived style through a dedicated late-fusion Dual-AdaLN pathway while preserving text-conditioned content structure. A cross-batch unpaired training protocol with latent cycle consistency enables joint learning across separate semantically rich and stylistically diverse datasets. As a proof-of-concept for controllable animation synthesis, we validate VISTA on rendered motion-capture references: it achieves the highest style recognition accuracy among video-conditioned methods while preserving competitive content alignment, and its decomposed 3-way classifier-free guidance provides independent, user-controllable calibration of the content-style balance at inference time.