How it works:

• Text Generation: Uses a Vision Language Model (VLM) to analyze each slide and generate natural narration
• Audio Synthesis: Uses Kokoro TTS to convert text to realistic speech
• Video Creation: Combines slide images with audio and smooth transitions using FFmpeg

Tips for best results:

• Provide context about your presentation to get more relevant narration
• Review and edit the generated notes to match your style
• Use the regeneration feature with specific instructions to refine individual slides
• Experiment with different voices and speeds to find what works best