[sdsacc]

On the product page there is an options for 2 months of email support. Does this include any help in designing grammars for applications? Or is it only for answering email questions?

[sdsacc]

Ah, I see. That makes sense. So my usage model is going to be recording a snippet using the iPhone mic and then passing the recording to runRecognitionOnWavFileAtPath. When I set up the AVAudioRecorder I am using the following setup:

// Define the recorder setting
NSMutableDictionary *recordSetting = [[NSMutableDictionary alloc] init];
[recordSetting setValue:[NSNumber numberWithInt:kAudioFormatLinearPCM] forKey:AVFormatIDKey];
[recordSetting setValue:[NSNumber numberWithFloat:16000.0] forKey:AVSampleRateKey];
[recordSetting setValue:[NSNumber numberWithInt: 1] forKey:AVNumberOfChannelsKey];

// Initiate and prepare the recorder
NSError *error = nil;
recorder = [[AVAudioRecorder alloc] initWithURL:outputFileURL settings:recordSetting error:&error];

Will this produce a file of the proper format for OpenEars or do I need to do some other conversion?

Thanks!

[sdsacc]

Thanks very much.

The docs did have some good examples but I am a little confused by what I am seeing. I have the following grammar defined:
NSDictionary *grammarOpening = @{
ThisWillBeSaidWithOptionalRepetitions: @[
@{ThisWillBeSaidOnce : @[
@{ ThisWillBeSaidOnce : @[@”TODAYS”]},
@{ ThisWillBeSaidOnce : @[@”DATE”]},
@{ ThisWillBeSaidOnce : @[@”IS”]}
]}
]};

Basically looking for the phrase “TODAYS DATE IS”. The issue is I get a hypothesis if I just say the word TODAY or I say the whole thing. This is true even if I group the words into a single line. Is that expected? If so how can I determine the whole phrase was spoken?

[Author]

Posts