FAQ

Recognizes anyone's voice. The recognition rate is affected somewhat by the clarity of pronunciation, the way you speak, the volume of your voice, etc.

It accommodates differences in intonation and speed of speech. However, dialects that are not included in the dictionary will need to be newly registered.

AmiVoice generates recognition results by looking at the words before and after the sentence, so stopping speech in the middle of a sentence can cause misrecognition.
Try to speak in whole sentences as much as possible.

Use the tips below.

• Please speak clearly and clearly
• Please speak in one sentence or punctuation mark. When multiple phrases are entered, Japanese context analysis works more effectively.
  Dictation analyzes all sounds coming from the microphone into Japanese sentences.
  Corrections and statements like "um" and "hmm" are considered words in context.

The recognition result is the word sequence with the highest probability calculated using the acoustic model and linguistic model, and recognition results for a mixture of kanji and kana are obtained directly from speech.
Speech recognition does not recognize hiragana and then convert it into kanji.

Recognition of hiragana characters one by one has not been put to practical use because many misrecognitions occur.