Index Ai Academy
Glossary · Phase 5 · Transformers

Signal-to-noise ratio

How much more power the wanted sound has than the noise, in decibels. At 0 the two are equally loud.

Try it now

Press Next. The lit arrow is the step described; the label under it is the output of that step for the word "mask".

Learn it properly

Recognition and synthesis as a pipeline · A whole speech recognizer for a made-up language: features, a guess per frame, collapsing repeats and blanks, a dictionary to repair mistakes, word error rate, and synthesis as the same stages run backward. About 40 minutes.

Other terms from the same lesson

  • CTC: A way to train and decode speech models with no marked boundaries: the model labels every frame with a sound or a blank, and repeats are merged before blanks are dropped.
  • lexicon: A list of words with the sounds each is made of.
  • phoneme: One of the units of sound a language's words are built from.
  • prosody: The rhythm, stress and melody of speech, which carry meaning the words alone don't spell out.
  • word error rate: Substitutions plus deletions plus insertions, divided by the number of words in the reference transcript.
Keep learning

More in plain English.

1286 terms in the glossary, each with the lesson that teaches it.

1286 terms

The glossary

Every word you'll meet in AI and machine learning, explained in everyday language.

Phase 5 · lesson 40 min

Recognition and synthesis as a pipeline

A whole speech recognizer for a made-up language: features, a guess per frame, collapsing repeats and blanks, a dictionary to repair mistakes, word error rate, and synthesis as the same stages run backward.

Never trained an AI? Teach a small one to write, right in your browser. It takes about 20 seconds and needs no account. Teach it to write