Building off of last week's progress, I have continued to look into the prospects of background noise cancellation as well as which would be the best method for me to employ in my own product. I have since determined one possible course of action for adding some degree of background noise cancellation to my project. As I have mentioned previously, the most ideal method for background noise cancellation involved both physical and software aspects. However, along with the development of separate software, this method would also require me to have two microphones. While it certainly isn't impossible for me to look into and develop the software side of this approach, I do not have access to two microphones to even make use of that software. So, instead of this approach, I have come to the conclusion that I can still employ some method of background noise cancellation that still utilizes physical and software aspects without needing two microphones. On the physical side, I will simply be using a single directional microphone. On the software side, I will be filtering the spectrogram using an intensity threshold under the assumption that the directional microphone will, at the very least, cause the recorded intensity of any background noise to decrease. I have begun implementing this strategy and plan on further tweaking and fine-tuning the software side going into the future.
Amol Kumar