How anpramila95/whiteboard-video works
This repository turns a topic, narration script, or existing SRT into a narrated whiteboard-animation video. The principal workflow is agent-guided script and narration preparation, timed subtitle/scene planning, scene illustration and region annotation, per-scene rendering, then video assembly and audio/subtitle finishing. The README documents several human/agent steps whose implementation is not present in the sampled source, so those links are documented workflow relationships rather than claims about inspected code.
Main entry points: run_video.py Video workflow · tts_narration.py Narration synthesis · parse_srt.py SRT scene planning · render_stream_whiteboard.py Scene renderer
Components 18
Script and timing
- Workflow skill
SKILL.md - Script to subtitles
scripts/script_to_srt.py - SRT scene planning
scripts/parse_srt.py - Deepgram segmentation
scripts/segment_deepgram.py - Annotation retiming
scripts/retime_annotations.py
Scene assets
- Scene image generation
scripts/batch_generate_images.py - Region annotations
- Annotation preview
scripts/render_annotation_preview.py
Whiteboard rendering
- Scene renderer
scripts/render_stream_whiteboard.py - Stroke renderer
scripts/stream_render.py
Video finishing
- Video workflow
run_video.py - Scene assembly
scripts/merge_scenes.py - Narrated video
Speech services
- Narration synthesis
scripts/tts_narration.py
Other components
- Creator / agent
- Edge TTS
- Vbee TTS
- Image generation API
Connections 22
- Creator / agent → Workflow skill: starts workflow
README.md - Workflow skill → Script to subtitles: creates draft
README.md - Script to subtitles → Narration synthesis: draft subtitles
README.md - Narration synthesis → Edge TTS: synthesizes with
README.md - Narration synthesis → Vbee TTS: synthesizes with
scripts/tts_narration.py - Narration synthesis → SRT scene planning: retimes subtitles
scripts/tts_narration.py - Creator / agent → SRT scene planning: provides SRT
README.md - SRT scene planning → Creator / agent: suggests scenes
README.md - Deepgram segmentation → Creator / agent: plans scenes
scripts/segment_deepgram.py - Creator / agent → Scene image generation: requests images
README.md - Scene image generation → Image generation API: requests images
scripts/batch_generate_images.py - Creator / agent → Region annotations: defines regions
README.md - Region annotations → Annotation preview: previews regions
README.md - Creator / agent → Scene renderer: renders scene
README.md - Scene renderer → Stroke renderer: uses stroke engine
scripts/render_stream_whiteboard.py - SRT scene planning → Annotation retiming: provides timing
scripts/retime_annotations.py - Creator / agent → Video workflow: runs workflow
README.md - Video workflow → Narration synthesis: calls synthesis
run_video.py - Video workflow → Scene renderer: calls renderer
run_video.py - Video workflow → Scene assembly: calls assembly
run_video.py - Scene assembly → Narrated video: joins scenes
scripts/merge_scenes.py - Video workflow → Narrated video: adds narration
README.md