The iPhone developers would be delighted to know that they can now use and contribute to the future of the iOS thanks to OpenEars. Though speech recognition has been a dream for quite some time, but iPhone developers can live that dream by being a part of the OpenEars. This particular software is an open source library for the iOS that executes text-to-speech as well as speech recognition for the iPad and iPhone. More about OpenEars OpenEars uses MITLM, CMU Flite and CMU Pocketsphinx libraries and this downloadable application can be installed in a Cocoa static library project. Once the iPhone developers have configured the project they would be able to target any of the architectures or the SDKs supported by the libraries. OpenEars is also executable in the Simulator, but outsource iPhone development team might find problems with the low latency audio driver since Simulator is not compatible with it. In order to tackle this issue, iPhone developers can utilize another Simulator compatible driver to debug the recognition logic. Nevertheless, it is best if you don’t run OpenEars on Simulator, and instead run it on the device itself. OpenEars also comes with downloadable instructions that iPhone developers can use to configure the library. The current version of this library is 0.912, but the instructions are for Xcode 4. If you are looking for Xcode 3 instructions, you can download the older 0.902 version of OpenEars. The instructions for this version are also meant for Xcode 4, which is known to be a more stable version. However, the latest version of this library does provide some advantages: What can OpenEars do? It offers iPhone developers all the functions that usually gobble up iOS CPU usage within just 8 percent of its total capacity. Bluetooth audio devices are supported It supports JSGF grammar and also generates dynamic ARPA language models OpenEars comes along with a new low-latency audio driver that boosts response speed You can now switch from one ARPA language model to another while multitasking It offers configurability of pitch, variance, speed and quality of voice, along with 8 preset voices for speech (male and female)
0 comments:
Post a Comment