In google speech recognition, when the conversion result of words and phrases or proper nouns is homophonic, such as character input, it does not change to the intended conversion candidate.

In Google Voice Input and Google Assistant, for example, "a famous person's name by kanji" and "a minor person's name" are always converted to a famous person's name, but even if Google Assistant registers only the minor person's name in Google's contact, it is not applicable.
The results were the same for personal computers and Android devices.

When I registered words and phrases in the Google Japanese input learning dictionary, the candidates for voice recognition did not change.

I tried using furigana in the "Minor Person's Name" section on Google, and the same reading as voice recognition in the memo section, katakana, hiragana, and half-width, but it didn't work.

I would like to change the priority of the personal conversion with Google voice input.
I would like to ask Google Assistant to be able to search for the same vocabulary (Chinese characters are different) by the name of the person registered in Google's contact, and to solve this problem if Google Assistant says, "What, email me" or "Call me."