Guys Jiawey and Jinda from ECE5760 class have build an adaptive noise canceler as their final project. Algorithm is based on least-mean-square (LMS) routine which is simple and efficient. Without going to deep in to theory of operation we can mention that algorithm is filtering noise contaminated signal in real time.
To do so, it needs two inputs, one for reference noise and another for signal which is contaminated with noise. Both inputs are fed in to DE2 board where algorithm filters noise out and produces cleaner sound to speakers.