Investigating optimizations in Android Speech Components & frameworks
- Text to speech engine
- OpenFst is a library for constructing, combining, optimizing, and searching weighted finite-state transducers (FSTs). FSTs have key applications in speech recognition and synthesis, machine translation, optical character recognition, pattern matching, string processing, etc.
- Speex is an Open Source/Free Software patent-free audio compression format designed for speech.
- the srec speech recognition engine
Desired Skills and Experience:
- Understanding of open source Audio frameworks (one or more from above list)
- Android audio framework in general.
- Handson experience in using AOSP sources to build and boot Android on hardware platforms.
- Profiling and benchmarking Android components to identify performance bottlenecks and improve the same.