How richard0403/dy-auto works
dy-auto automates short-video creation and Douyin publishing. Its main workflow starts in the CLI or desktop UI, prepares text and images, synthesizes narration, renders and subtitles a video, then optionally uploads it. The repository also documents batch/random content generation and includes an optional GAN-based lip-sync/virtual-person path. The graph follows the sampled orchestration and renderer code; documented integrations such as Whisper and FFmpeg are included, but unsampled implementation details are not inferred.
Main entry points: speech_video_gen.py Speech workflow · video_gen.py Video renderer · speech_control.py Upload control · main_win_paddle_gan.py GAN desktop mode
Components 21
Content creation
- CLI entry
dy-auto/main.py - Desktop UI
dy-auto/main_win.py - Text topics
DyText - Image library
DyPic - Speech workflow
dy-auto/speech_video_gen.py
Media pipeline
- Media preparation
dy-auto/file_handle.py - Voice synthesis
dy-auto/speech_voice_gen.py - Narration audio
DyTemp/temp_voice - Video renderer
dy-auto/video_gen.py - Speech transcription
dy-auto/speech_srt_gen.py - Subtitle correction
dy-auto/srt_correct.py - Rendered video
DyTemp/temp_video
Publishing
- Upload control
dy-auto/speech_control.py - Douyin creator
Optional video modes
- GAN desktop mode
dy-auto/main_win_paddle_gan.py - GAN transforms
dy-auto/PaddleTools/GAN.py - GAN settings
dy-auto/PaddleTools/config.py
Other components
- User
- Azure Speech
- FFmpeg
- Whisper
Connections 26
- User → CLI entry: starts
dy-auto/main.py - CLI entry → Speech workflow: runs
dy-auto/main.py - User → Desktop UI: enters content
dy-auto/main_win.py - Desktop UI → Speech workflow: generates video
dy-auto/main_win.py - Speech workflow → Text topics: selects topics
dy-auto/speech_video_gen.py - Speech workflow → Image library: selects images
dy-auto/speech_video_gen.py - Speech workflow → Voice synthesis: requests narration
dy-auto/speech_video_gen.py - Voice synthesis → Azure Speech: synthesizes speech
dy-auto/speech_voice_gen.py - Voice synthesis → Narration audio: writes audio
dy-auto/speech_voice_gen.py - Speech workflow → Media preparation: prepares images
dy-auto/speech_video_gen.py - Speech workflow → Video renderer: builds video
dy-auto/speech_video_gen.py - Video renderer → Narration audio: uses narration
dy-auto/video_gen.py - Video renderer → FFmpeg: renders and overlays
dy-auto/video_gen.py - Video renderer → Speech transcription: requests subtitles
dy-auto/video_gen.py - Speech transcription → Whisper: transcribes audio
readme.md - Video renderer → Subtitle correction: corrects captions
dy-auto/video_gen.py - Video renderer → Rendered video: writes result
dy-auto/video_gen.py - CLI entry → Upload control: runs publishing flow
dy-auto/main.py - Upload control → Speech workflow: generates upload video
dy-auto/speech_control.py - Upload control → Douyin creator: uploads video
dy-auto/speech_control.py - Desktop UI → Upload control: uploads when enabled
dy-auto/main_win.py - Upload control → Rendered video: uploads file
dy-auto/speech_control.py - User → GAN desktop mode: starts optional mode
dy-auto/main_win_paddle_gan.py - GAN desktop mode → GAN settings: loads settings
dy-auto/main_win_paddle_gan.py - GAN desktop mode → GAN transforms: runs transforms
dy-auto/main_win_paddle_gan.py - GAN desktop mode → Upload control: optionally uploads
dy-auto/main_win_paddle_gan.py