Generate a video using HeyGen API.

Video Inputs (1-50 scenes):

• character: Avatar or Talking Photo settings (optional)
• voice: Text, Audio, or Silence voice settings (required)
• background: Color, Image, or Video background (optional)

Character Types:

• Avatar: avatar_id, scale (0-5), avatar_style (circle/normal/closeUp), offset, matting
• Talking Photo: talking_photo_id, scale (0-2), talking_photo_style, talking_style, expression

Voice Types:

• Text: voice_id, input_text, speed (0.5-1.5), pitch (-50 to 50), emotion, locale
• Audio: audio_url or audio_asset_id
• Silence: duration (1-100 seconds)

Background Types:

• Color: hex value (e.g., "#f6f6fc")
• Image: url or image_asset_id, fit (cover/crop/contain/none)
• Video: url or video_asset_id, play_style, fit

Additional Options:

• dimension: Video dimensions (default: 1280x720)
• title: Video title
• caption: Add captions (text voice only)
• callback_id: Custom callback ID
• callback_url: Dynamic callback URL
• folder_id: Output folder
• test: Test mode

Returns: video_id: The ID of the generated video task