Audio Processing

4. Audio Models

• Audio generation models produce audio from text. Voiceover for videos. Services for hard of hearing. Phone automations. A school finds that students pay better attention when they have audio to go with the content. The students get to pick which voice is most appealing to them. They use llms to adapt the content level easier or more complex. Together the audio and content help students improve understanding and retention.
• Audio Transcription models. A call center has its workers fill out a form after each call to keep statistics on the call type the solution and any mistakes made. They often forget what they said and it takes a long time for them to listen back to a recording. Now the audio is automatically transcribed into text so it can be automatically used to classify the calls or they can manually read back what they said and fill out the form.
• Two businesses are having a discussion regarding a new partnership. They are talking in person and discussing the details. Their conversation is transcribed so that they can review what they said later and have proof of past discussion details.