Home › Forums › OpenEars › Increasing speech recognition accuracy using a custom languagemodel? › Reply To: Increasing speech recognition accuracy using a custom languagemodel?
Does it work when you compile for a device? I checked the demo version of SaveThatWave and it definitely has that architecture in it, so the possibilities are that that -ObjC linker flag isn’t added to the target that is being run, or the entry for the target in the Framework Search Paths build setting that was automatically created at the time of adding the framework is incorrect or has been mangled by Xcode (this is an occasional Xcode bug), or the framework wasn’t added to the target that is being run. Otherwise there’s nothing that should be preventing it from working. Since it fails at runtime rather than compile time, I’d be suspicious of the linker flag.