See Before You Code: Learning Visual Priors for Spatially Aware Educational Animation Generation

TL;DR

This paper introduces OmniManim, a framework for generating educational animations from natural language that ensures visual quality by incorporating render-feedback and explicit visual planning.

Contribution

It presents a novel render-feedback-aware code generation framework with visual planning, datasets, and evaluation protocols for educational animation quality.

Findings

OmniManim improves render quality over baselines on EduRequire-500.

Explicit visual planning components are crucial for quality improvements.

Constructed datasets enable comprehensive evaluation of animation generation methods.

Abstract

Large language models can generate executable code for educational animations, but the resulting renders often exhibit visual defects, including element overlap, misalignment, and broken animation continuity. These defects cannot be reliably detected from the code alone and become apparent only after execution. We formalize this problem as render-feedback-aware constrained code generation: given a natural language specification, the model must generate executable code whose rendered output satisfies structured quality criteria that can be evaluated only after rendering. To address this problem, we introduce OmniManim, a render-feedback-aware educational animation generation framework built around a shared scene state, explicit visual planning, structured post-render diagnostics, and localized repair. Within OmniManim, the Vision Agent is a task-specific visual planning module: it predicts sparse keyframe layouts with coarse-to-fine bounding-box denoising and optimizes an interpolation-aware objective to reduce intermediate-frame failures induced by downstream animation interpolation. We further construct two datasets, ManimLayout-1K and EduRequire-500, and provide a reproducible evaluation protocol covering executability, instructional quality, visual quality, and efficiency. On EduRequire-500, OmniManim improves measured render quality over both single-model baselines and existing multi-agent frameworks. Systematic ablation studies further verify that explicit visual planning, especially its coarse spatial prior, bounding-box refinement, and interpolation-aware optimization, is central to these gains.