===Adventure Sphinx 0.1=== This package should build and install the exectuable called adventure_sphinx, which is an entirely speech based role playing adventure game. The game itself is "Colossal Cave Adventure", and the speech recognition engine used is CMU Sphinx, and the TTS engine used is Festival. PortAudio is used for recording input. You will need to have these libraries installed to build this program. (They are in a nonstandard place, you might have to tell autoconf where they are). Once installed, simply execute "adventure_sphinx" to start playing the game. The only command line options are -sthresh, which determines the threshold in dB which your voice must exceed to trigger the recording start, and -ethresh, which similarly determines at what level it will be assumed you have stopped speaking (Use negative numbers, as you want to measure dB *below* the maximum level of your input. Also, the recording bars will visually indicate where these thresholds are, so you can tweak the parameters to match your current environment). Any other option will be treated as a save file from which to load the game (as per the original "adventure" game). Once the game has started, all interaction is voice based, except if you are asked to enter the name of a save file. The text output is just for debugging, and it can help you understand how the speech recognizer is misunderstanding you. Have fun playing, and remember, it really can't understand even moderately complex sentences. You sentences essentially have to all be two words, ignoring filler words like prepositions. Also, if it fails to recognize a word, try again, but if after several tries it's still not working, most likely that word is not in its vocabulary. Credits: This program is derived from a project for a class in "Spoken Language Processing" written by Jeremy Salwen, Connie Liu, and Laima Tazmin. This is a rewritten version by Jeremy Salwen. Contact: If you have comments, suggestions, bug reports, or hate mail, please email jeremysalwen@gmail.com.