A maple-sugar shack wants NVIDIA Riva ASR to recognize local terms such as “spile,” “niter,” and “hydrometer.” Staff prepare a word list plus example utterances. How should that material be treated for Riva ASR customization?
Select an answer to reveal the explanation.
Short Explanation
Think of the shack’s jargon as a mini dictionary with audio examples attached. Riva ASR customization eats that speech-plus-text asset—not a model bake-off or an image album. Get the words and utterances into the customization set first.
Full Explanation
Site-specific terms become useful when they are packaged as speech and text that an ASR customizer can learn from. A word list with example utterances is that customization corpus. Comparing recognizer architectures, building CLIP pairs, or editing Helm charts address different problems and do not substitute for the vocabulary dataset.