How yenloned/aistoryvideogenerator works
AI Story Video Generator turns a keyword or supplied story into a narrated, illustrated video. The principal workflow is initiated through main.py: it obtains structured script data, generates scene images and paragraph audio, then composes them with synchronized subtitles into an MP4. Interactive topic selection reads data/topics.json, and batch generation repeatedly invokes the same pipeline. README capabilities include local Ollama script generation, Stable Diffusion imagery, Coqui/Piper narration, subtitle generation and FFmpeg composition. Source excerpts are partial, so links for unsampled implementation details are limited to documented behavior.
Main entry points: main.py Video pipeline · generate_video.py Video composition · generate_audio.py Audio generation · generate_images.py Image generation
Components 18
Entry and orchestration
- Batch runner
batch_generate.py - Video pipeline
main.py - Topic catalog
data/topics.json
Story preparation
- Script generation
scripts/generate_script.py - Story conversion
main.py - Script JSON
Scene assets
- Image generation
scripts/generate_images.py - Audio generation
scripts/generate_audio.py - Scene images
- Narration audio
Video composition
- Video composition
scripts/generate_video.py - Subtitled MP4
Other components
- User
- Ollama Qwen
- Stable Diffusion
- Coqui TTS
- Piper TTS
- FFmpeg
Connections 21
- User → Video pipeline: provides input
- User → Batch runner: provides keywords
- Batch runner → Video pipeline: runs per keyword
- Video pipeline → Topic catalog: reads topics
- Video pipeline → Story conversion: converts supplied story
- Video pipeline → Script generation: requests script
- Script generation → Ollama Qwen: requests generation
- Video pipeline → Script JSON: writes script
- Video pipeline → Image generation: requests images
- Image generation → Stable Diffusion: generates scenes
- Image generation → Scene images: writes images
- Video pipeline → Audio generation: requests narration
- Audio generation → Coqui TTS: synthesizes speech
- Audio generation → Piper TTS: synthesizes speech
- Audio generation → Narration audio: writes audio
- Video pipeline → Video composition: requests composition
- Video composition → FFmpeg: assembles media
- Video composition → Script JSON: reads script
- Video composition → Scene images: reads images
- Video composition → Narration audio: reads audio
- Video composition → Subtitled MP4: writes subtitled video