Is there a way to find the individualword when OpenEars recognizes them in loop.

[skp]

Hi,
I am planning to use Open Ears for my app. I have the predefined file(.arpa, .dic) added to my project. Speech detection works fine.

The issue I have is that multiple words are recognized during the speech and are clubbed into one word “hypothesis” in the method – (void) pocketsphinxDidReceiveHypothesis:(NSString *)hypothesis recognitionScore:(NSString *)recognitionScore utteranceID:(NSString *)utteranceID .

I use this as the detected word.

Is there a way to find the words as and when Open ears recognizes them in loop.

Eg: My file has these 3 words

HI
YELLOW FRUIT
WHEN

When open ears starts to listen in loop if user speaks fast enough to say all these words, the received hypothesis word shows as one word “HI YELLOW FRUIT WHEN”

Instead, I want to know the words individually when speech is detected.

[Halle Winkler]

Welcome,

Is the question about how to divide an NSString into its component words?

[skp]

Hi,
I am already dividing the string into the predefined words. However, wanted to know if Open Ears has any method of showing the identified word as and when it detects it rather than updating it as a phrase in “hypothesis”

[Halle Winkler]

The plugin RapidEars can show recognition results in real time (that is, live as the user is speaking). There’s more info here: https://www.politepix.com/rapidears

[Author]

Posts