Kvíz spojený s průvodcem · Střední úroveň

Video-to-Audio Generation a AI Foley Quiz

Otestujte si, co víte o modelech video-audio, jak udržují zvuk v synchronizaci s obrazem a kde AI Foley stále zaostává.

Související vodicí cestyVideo To Audio Generation A AI Foley
Otázka 1 z 8

What most distinguishes video-to-audio models from text-to-sound-effect tools?

Náhled každé otázky8 otázky

Odpovědi a vysvětlení se zobrazují, jak budete v kvízu.

  1. What most distinguishes video-to-audio models from text-to-sound-effect tools?
    • They only generate music
    • They use the footage itself to match sound to on-screen events and timing
    • They require recorded microphone input
    • They cannot accept any text prompt
  2. The film craft of Foley is named after whom?
    • Jack Foley of Universal
    • A sound effects software company
    • A microphone brand
    • A film festival
  3. Which system generated video and audio together rather than adding sound afterwards?
    • VGGSound
    • Synchformer
    • KLIP
    • Google's Veo 3
  4. What role does an optional text prompt play in a V2A model?
    • It replaces the video entirely
    • It sets the video frame rate
    • It adds details pixels cannot supply, such as the surface material
    • It translates dialogue
  5. Why might a V2A model make a plastic prop sound metallic?
    • It cannot reliably infer material from appearance alone
    • The video codec distorts sound
    • Plastic cannot be recorded
    • The model only produces metal sounds
  6. What is VGGSound?
    • A mixing plugin
    • A synchronization metric
    • A dataset built from YouTube video clips with sound
    • A Google audio model
  7. What does MMAudio pair with lower-rate semantic image features to improve timing?
    • Lyrics transcriptions
    • Higher-rate synchronization features from a Synchformer model
    • MIDI note data
    • Pouze náhodný hluk
  8. How is V2A output best treated in professional sound work?
    • As a final mix ready for release
    • As a replacement for all recorded effects
    • As unusable
    • As a first pass that a sound editor refines