Speech and Face Recognition
Over Thanksgiving 2009, I embarked on Crazy Weekend Project #2. With the goal of exploring technologies that would be of use in an unattended informational display screen, I looked at the System.Speech.Recognition namespace in .Net 3.5, as well as facial detection and recognition features in OpenCV. I found that the speech recognition features were quite easy to use, just a few lines of code and you have an application that can respond to voice commands. On the facial detection side, I had less immediate success, fighting with stupid library issues more than theoretical or application concerns. In the end, however, I was able to write an application that was able to recognize characters from an episode of Star Trek: The Next Generation with a high degree of success. I used this technology strictly for good, and applied what I found to remove Wesley Crusher from the episode.
You can read about my failures and successes here: http://www.mathpirate.net/log/category/crazy-weekend-project-2/