← CHI 2026 map

CHI '26 · Honorable mention · full-paper review · confidence medium-high

Notational Animating: An Interactive Approach to Creating and Editing Animation Keyframes

Xinyu Shi , Li-Yi Wei , Nanxuan Zhao , Jian Zhao , Rubaiat Habib Kazi

DOI PDF Program page Page JSON

This is a thoughtful CHI paper that reframes animator sketches as ambiguous, high-level motion cues and then builds an interaction loop around that insight. The novelty is strongest in the structured source-path-target representation plus editable feedback, while the evidence is credible but still preliminary.

Axes Lens

Rare contribution shape, typical evidence profile. The point here is not a score. It is to show what kind of claim the paper makes, and whether the evidence pattern is unusual or baseline in this 268 -review set.

Review Summary

This paper’s main contribution is not simply that it uses AI to generate animation keyframes from sketches; that idea would be too familiar. The more interesting move is conceptual: it treats animator notations as contextual, ambiguous, and combinational artifacts that should be interpreted rather than obeyed literally. That reframing is then operationalized through a structured representation of motion intent, expressed as source, path, and target, and through an interactive system that lets users refine the interpretation with dynamic widgets and a closed feedback loop. In CHI terms, that makes the paper more than a tool demo: it proposes a design framework for turning sketch-based motion authoring into a controllable human-AI interaction process. The evidence packet supports this reading well. The abstract explicitly states the ambiguity problem, the formalization step, and the feedback loop, and the findings indicate that professional animators found the interaction intuitive and useful. At the same time, the validation is clearly bounded. The study is preliminary, qualitative, and based on professional animators rather than a broader population; the tasks are described as toy-scale relative to production work; and the discussion acknowledges open issues around model fidelity, latency, and the risk of unsolicited creative intent. So the paper’s claim is credible as a framework and interaction technique for a specific task class, but not yet as a general solution for animation authoring. I would read it as a solid honorable-mention-level contribution: conceptually sharp, well-positioned, and grounded in a plausible prototype, with validation that is sufficient for early-stage CHI evidence but not enough to support broad deployment claims.

What Changed

Canon before

Animator sketches are typically treated as direct, local drawing instructions or informal annotations; this paper reframes them as higher-level, ambiguous motion cues that need structured interpretation and iterative refinement.

Departure from common sense

The paper argues against treating animator sketches as fixed, isolated constraints. Instead, it positions them as contextual and combinational cues whose meaning depends on interpretation and feedback, which is a notable shift from a literal sketch-to-motion assumption.

Actual novelty

The core novelty is the formalization of free-form sketched notations into a structured animation representation and the use of that representation inside an interactive authoring loop with editable widgets and ambiguity resolution. That combination turns sketch interpretation into a controllable authoring workflow rather than a one-shot generation step.

Evidence

The paper grounds its claims in a design study, a system proposal, and a preliminary qualitative evaluation with professional animators. Evidence supports the reframing of sketches as ambiguous motion cues, the structured source-path-target representation, and user reports that the interaction was intuitive and useful. Validation is promising but limited in scale and scope.

Limits

Method limits

The evaluation is qualitative and preliminary, with a small expert sample and toy-scale tasks rather than production workflows. The paper also leaves open how well the approach generalizes beyond the studied cases and how robust the interpretation pipeline is under broader ambiguity.

Deployment limits

Practical deployment depends on model fidelity, latency, and the ability to preserve or infer intended motion without introducing unwanted creative changes. The system is also constrained by the need for iterative user correction when ambiguity cannot be resolved automatically.

Boundary conditions

The approach is most plausible when users can express motion intent through sketched notations and are willing to refine outputs through a feedback loop. It is less certain for novice users, long-horizon production settings, or cases requiring precise control over complex 3D motion and occlusion.

Manual check: No manual check requested from the evidence packet, but the exact section offsets in the source spans are approximate and the validation remains preliminary.

Position in field

This work sits at the intersection of animation authoring, sketch-based interaction, and AI-assisted creative tools. Its contribution is best read as a structured interaction framework for translating ambiguous animator notation into editable motion intent, rather than as a fully automated animation generator.

Abstract

We introduce the concept of notational animating, an interaction paradigm for animation authoring where users sketch high-level notations over static drawings to indicate intended motions, which are then interpreted by automatic methods (e.g., GenAI models) to generate animation keyframes. Sketched notations have long served as cognitive instruments for animators, capturing forces, poses, dynamics, paths, and other animation features. However, such notations are often contextual, ambiguous, and combinational based on our analysis of 135 real-world sketches. To facilitate interpretation, we first formalize these notations into a structured animation representation (i.e., source, path, and target). We then built an animation authoring system that translates high-level notations into the formalized intended animation, provides dynamic UI widgets for fine-grained parameter control, and establishes a closed feedback loop to resolve ambiguity. Finally, through a preliminary study with animators, we assess the usability of notational animating, reflect its affordance, and identify its contexts of use.