Bài kiểm tra nhân bản giọng nói

Câu trả lời và giải thích xuất hiện khi bạn làm bài kiểm tra.

1. A speaker permits a synthetic voice for one educational lesson. What does that permission cover?
   • Only the defined scope unless broader use is separately authorized
   • Every future advertisement and message
   • Any use of recordings found online
   • All uses by anyone who downloads the model
2. A person’s recording is publicly available online. What does that establish about cloning rights?
   • The person consented to every use
   • Availability alone does not establish permission
   • The voice may be used in any language
   • No disclosure is needed
3. A clone resembles the speaker. Which additional set of speech qualities does the guide recommend evaluating?
   • The model’s training company
   • Only resemblance
   • Pronunciation, intelligibility, emotional tone, and stability
   • Whether the audio is compressed
4. Why can a good-sounding clone still misrepresent a person?
   • Similarity verifies the script
   • It always includes the person’s live approval
   • It cannot generate new words
   • It may say content the person never approved or imply a meaning they did not intend
5. How should access to source recordings and voice models be managed?
   • Store reusable assets in broadly accessible folders
   • Restrict who can generate outputs and keep an appropriate use record
   • Share model artifacts publicly by default
   • Remove all records of use
6. When should synthetic speech be disclosed?
   • Never for narration
   • Only when it sounds artificial
   • When listeners might mistake it for an original recording
   • Only after a complaint
7. A voice clone is used for an urgent payment request. How should the recipient verify identity?
   • Use an established independent channel rather than voice similarity
   • Call a number supplied in the message
   • Trust the familiar voice
   • Ask for more generated audio
8. What should an authorization record specify?
   • The clone’s emotional tone
   • Only the source file format
   • The number of times a model can run
   • Who may generate speech, for which purposes, and how approval can change or be withdrawn