Open-Source Workflow Creates Talking Slide Avatars for Teaching
A new open-source workflow enables instructors to generate talking slide avatars by combining OpenVoice for text-to-speech and voice cloning with Ditto-TalkingHead for audio-driven talking-image synthesis. This approach addresses the loss of instructor presence in online, hybrid, and asynchronous slide-based teaching by transforming a script and static portrait into a short narrated video. The study frames this as a pedagogical and production solution rather than a purely technical one.
Key facts
- Workflow integrates OpenVoice and Ditto-TalkingHead
- OpenVoice handles text-to-speech and voice cloning
- Ditto-TalkingHead performs audio-driven talking-image synthesis
- Output is a short narrated video embedded in slide decks or HTML materials
- Addresses loss of instructor presence in online teaching
- Aims to restore narrative continuity and expressive framing
- Practice-based analysis of an open-source workflow
- Targets higher education slide-based teaching
Entities
—