Jonathan Windle

Dr Jonathan Windle · Research Scientist, Epic Games, London

Voice in.Motion out.

I build AI that turns speech into animation for digital humans: faces, tongues, heads and hands.

Eight animation frames of a figure gesturing, driven by a speech waveform below, with animation curves linking each frame to the audio
frames · rig curves · audiospeech-driven, frame by frame

IN PRODUCTION · SIGGRAPH 2025

EpicADA

Controllable and Expressive Audio-Driven Animation

Full face, tongue and head animation straight from speech, written directly to MetaHuman rig controls, with emotion detected from the voice or chosen by the animator. It ships in Unreal Engine as MetaHuman audio-driven animation and is still being improved.

S. Taylor, S. Medina, J. Windle, E. Saez, I. Matthews

Read the paper Summary

Research

All papers on Google Scholar →

Earlier: CLICKA: identity cues from keystroke dynamics, Computers & Security, 2022.

Path

  1. Research Scientist
    Epic Games · since July 2024
  2. PhD, Digital Humans: Automatic Character Animation
    University of East Anglia · 2024
  3. Research Intern
    Epic Games
  4. Research Associate, keystroke dynamics
    University of East Anglia
  5. Software Developer Intern
    Boeing Defence UK · 2017–18

Recently

Working on digital humans?

I'm always happy to talk speech-driven animation, research collaborations and speaking.

Message me on LinkedIn Scholar GitHub ORCID