A weathervane loft wants NVIDIA Riva TTS to read care notices aloud. A volunteer points at a pile of vane photographs as “the TTS data.” What should the team request instead?
Select an answer to reveal the explanation.
Short Explanation
TTS is a reader, not a photo gallery. Feed it the words you want spoken—and voice samples if you care how they sound—not a stack of vane pictures.
Full Explanation
A TTS corpus is text written in spoken style, optionally with target-voice audio samples. Image albums and CLIP pairs support vision–language work, not Riva TTS. Architecture redesign does not turn stills into speakable scripts.