For doing accurate speech recognition on Android, Google’s speech recognizer, along with the android speech api is readily available. However, the problem is that this API require you to install the Google search on to the phone, and have a active data connection (speech database and the recognition is done in the backend).
There is an open source project called CMUSphinx, which has a mobile edition and has an Android demo. However most of the instructions found online on how to build didn’t really work very well for me. The next few posts are to document my step by step process for future use.