Agentic video generation framework that turns a story idea, screenplay, or novel into a multi-shot video by coordinating script, storyboard, character, and video generation agents.
geeklee/srt-whiteboard-animation
An AI skill that turns SRT subtitles into hand-drawn whiteboard animation videos on a warm beige paper background, exported as MP4.
- Stars
- 4,616
- Forks
- 720
- Open issues
- 8
- Last commit
- 3 months ago
What is it?
This project is an AI agent skill that converts an SRT subtitle file into a whiteboard-style hand-drawn animation. Each scene is built from a line drawing and an annotation file that defines regions, drawing order, timing and the subtitle line each element belongs to. Elements appear in narrative order, and within each region the pen draws continuous strokes: first the ink outline, then color fill.
The workflow is step by step. The skill parses the SRT and suggests scene splits of roughly 25 to 35 seconds, then generates a line drawing, creates annotations, and lets the user adjust regions and timing in a browser-based preview page before rendering. Single scenes are rendered to MP4 and can be merged into a longer video. Output uses a minimal style with a warm paper background, dark gray sketch lines, and sparing red, orange and blue accents.
The project is written in Python and uses OpenCV, with a helper script that prepares an isolated virtual environment.
Key features
- 01Parses SRT subtitles and suggests scene splits of 25 to 35 seconds.
- 02Draws each region as a continuous stroke, first ink outline then color fill.
- 03Manages regions, timing, subtitle links and protected overlap areas in a JSON annotation file.
- 04Includes a browser preview page for adjusting regions, order and timing.
- 05Renders single scenes to MP4 and merges multiple scenes into one video.
Use cases
- Turning educational explanations or course subtitles into hand-drawn whiteboard videos.
- Producing narrated story or short-video content in a simple sketch style.
- Creating knowledge-sharing clips where each scene follows the subtitle sequence.
Quick start
- 1
Check the Python environment setup
python scripts/prepare_env.py --check - 2
Prepare the isolated virtual environment
python scripts/prepare_env.py - 3
Parse the SRT subtitles into suggested scenes
python scripts/parse_srt.py <字幕.srt> --target-sec 30 --min-sec 25 --max-sec 35 - 4
Render a scene to MP4 with the stream renderer
<ENV_PY> scripts/render_stream_whiteboard.py <图片路径> <标注路径> <输出.mp4> assets/drawing-hand.png --ink-path grid --color-fill contour-wipe
Things to know
- Each scene needs a hand-made annotation file, so the workflow is labor-intensive.
- The process requires confirmation at several stages before rendering, so it is not fully automatic.
- The project is written as an AI agent skill, so its workflow assumes an AI assistant.
Topics
FAQ
- What is srt-whiteboard-animation?
- An AI skill that turns SRT subtitles into hand-drawn whiteboard animation videos on a warm beige paper background, exported as MP4. This project is an AI agent skill that converts an SRT subtitle file into a whiteboard-style hand-drawn animation. Each scene is built from a line drawing and an annotation file that defines regions, drawing order, timing and the subtitle line each element belongs to. Elements appear in narrative order, and within each region the pen draws continuous strokes: first the ink outline, then color fill.
- Is srt-whiteboard-animation free to use? What license does it have?
- srt-whiteboard-animation is open source under the MIT license. Check the license terms before using it in commercial products.
- Is srt-whiteboard-animation still maintained?
- Updated within 6 months. The last commit to srt-whiteboard-animation was on Jul 27, 2026.
- What can I use srt-whiteboard-animation for?
- Turning educational explanations or course subtitles into hand-drawn whiteboard videos. Producing narrated story or short-video content in a simple sketch style. Creating knowledge-sharing clips where each scene follows the subtitle sequence.
- How do I install srt-whiteboard-animation?
- Check the Python environment setup: python scripts/prepare_env.py --check Prepare the isolated virtual environment: python scripts/prepare_env.py Parse the SRT subtitles into suggested scenes: python scripts/parse_srt.py <字幕.srt> --target-sec 30 --min-sec 25 --max-sec 35 Render a scene to MP4 with the stream renderer: <ENV_PY> scripts/render_stream_whiteboard.py <图片路径> <标注路径> <输出.mp4> assets/drawing-hand.png --ink-path grid --color-fill contour-wipe
- Where can I get srt-whiteboard-animation?
- The source code and installation instructions are on GitHub: https://github.com/geeklee/srt-whiteboard-animation
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A benchmark suite and evaluation toolkit that scores text-to-video and image-to-video generative models across many quality and semantic dimensions.
A Codex and Claude Code skill that turns a topic, article or story into a narrated hand-drawn explainer video, with stroke-by-stroke drawing and Chinese handwriting.
A Codex Skill that turns a Chinese spoken line or abstract idea into pen-line sketch explainer animation assets with selective colored-pencil fill, generated via Gemini.
An Agent Skill that turns a topic into a narrated explainer video with voiceover, hard subtitles and cover images, rendering HTML frames locally to MP4.
Agent memory system that stores, recalls and reflects on memories so AI agents learn over time instead of only replaying chat history.