What is the usual input to an end-to-end speech synthesis system before it generates spoken output?

In an end-to-end speech synthesis system, the model can generate audio directly from text features.

Input for End-to-End Speech Generation

Match each part of a speech synthesis system to its function.

Order the steps in a direct text-to-speech pipeline from input to output.

Why is end-to-end text-to-speech grouped with direct learning of rich outputs?

A text-to-speech system learns to turn speech recordings into written transcripts.

Rich Output in an End-to-End Speech Synthesizer

Match each term to its role in a direct speech-synthesis system.

Order the reasoning steps for identifying end-to-end speech synthesis as a directly learned rich-output task.

How an End-to-End Speech Model Learns a Rich Output

Design an end-to-end speech output pipeline.

What does an end-to-end speech synthesis system take in and produce?