Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Sketch Animation: State-of-the-art Report

Gaurav Rai, Ojaswa Sharma

TL;DR

Sketch animation addresses transforming abstract hand-drawn sketches into dynamic visual narratives by integrating a spectrum of techniques from keyframe interpolation to data-driven and diffusion-based methods. The paper surveys traditional, physics-based, mocap/data-driven, and learning-based approaches, and analyzes recent trends in generative AI, cross-domain motion transfer, and multimodal control. It introduces a taxonomy, reviews representative datasets and evaluation metrics, and discusses limitations such as temporal coherence, data scarcity, and real-time inference. The findings underscore the potential of multimodal, interactive pipelines—equipped with physics reasoning and diffusion priors—to accelerate production and enable accessible, semantically controllable animation in education, entertainment, and AR/VR.

Abstract

Sketch animation has emerged as a transformative technology, bridging art and science to create dynamic visual narratives across various fields such as entertainment, education, healthcare, and virtual reality. This survey explores recent trends and innovations in sketch animation, with a focus on methods that have advanced the state of the art. The paper categorizes and evaluates key methodologies, including keyframe interpolation, physics-based animation, data-driven, motion capture, and deep learning approaches. We examine the integration of artificial intelligence, real-time rendering, and cloud-based solutions, highlighting their impact on enhancing realism, scalability, and interactivity. Additionally, the survey delves into the challenges of computational complexity, scalability, and user-friendly interfaces, as well as emerging opportunities within metaverse applications and human-machine interaction. By synthesizing insights from a wide array of research, this survey aims to provide a comprehensive understanding of the current landscape and future directions of sketch animation, serving as a resource for both academics and industry professionals seeking to innovate in this dynamic field.

Sketch Animation: State-of-the-art Report

TL;DR

Sketch animation addresses transforming abstract hand-drawn sketches into dynamic visual narratives by integrating a spectrum of techniques from keyframe interpolation to data-driven and diffusion-based methods. The paper surveys traditional, physics-based, mocap/data-driven, and learning-based approaches, and analyzes recent trends in generative AI, cross-domain motion transfer, and multimodal control. It introduces a taxonomy, reviews representative datasets and evaluation metrics, and discusses limitations such as temporal coherence, data scarcity, and real-time inference. The findings underscore the potential of multimodal, interactive pipelines—equipped with physics reasoning and diffusion priors—to accelerate production and enable accessible, semantically controllable animation in education, entertainment, and AR/VR.

Abstract

Sketch animation has emerged as a transformative technology, bridging art and science to create dynamic visual narratives across various fields such as entertainment, education, healthcare, and virtual reality. This survey explores recent trends and innovations in sketch animation, with a focus on methods that have advanced the state of the art. The paper categorizes and evaluates key methodologies, including keyframe interpolation, physics-based animation, data-driven, motion capture, and deep learning approaches. We examine the integration of artificial intelligence, real-time rendering, and cloud-based solutions, highlighting their impact on enhancing realism, scalability, and interactivity. Additionally, the survey delves into the challenges of computational complexity, scalability, and user-friendly interfaces, as well as emerging opportunities within metaverse applications and human-machine interaction. By synthesizing insights from a wide array of research, this survey aims to provide a comprehensive understanding of the current landscape and future directions of sketch animation, serving as a resource for both academics and industry professionals seeking to innovate in this dynamic field.

Paper Structure

This paper contains 32 sections, 2 equations, 26 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Overview
  3. Contribution
  4. Structure of the review
  5. A brief history of animation techniques
  6. Background
  7. Sketch fundamentals
  8. Sketch representations
  9. Inbetweening approaches for sketch animation
  10. Sketch keyframe interpolation
  11. Keyframe-free approaches
  12. Physics-based sketch animation
  13. Data-driven and motion capture techniques
  14. 2D sketch to 3D animation
  15. Learning based methods
  16. ...and 17 more sections

Figures (26)

  • Figure 1: Number of papers published over the years on sketch animation.
  • Figure 2: Distribution of publications within sketch animation across venues over the time span of 2001-2025.
  • Figure 3: Number of papers published from 2001–2025 on 2D/3D, raster/vector, automatic/manual, and real-time/offline.
  • Figure 4: Sketch animation flowchart.
  • Figure 5: Raster and vector representations of the given sketches. Images adopted from rai2024sketchanimli_sketch-r2cnn_2018rai2024enhancing.
  • ...and 21 more figures