I'm hopig that noise cancelling will improve to learn how to cancel voices, so that when hearing a coworker's grating voice or loud laughing you can prss a button "remember voice" and this will be recorded for future cancellation.
Noise cancellation is designed to cut out the ambient noise - which is mostly white - of which there is a lot in a big city. It does this with a phase cancellation of what is probably a fairly predictable waveform. Human voices I can imagine are not that predictable. Maybe advances in machine learnings and processing may be able to provide better cancellation techniques.
Noise cancelling cannot cancel out voices. It's designed to cancel out predictable lower frequency sounds such as fans, A/C system fans, engine rumbles and jet engine noise.
The noise cancelling system has an inherent delay in processing the sound it's listening to output an opposing waveform. Since voices are not predictable, outputting a waveform after a slight delay will just cause it to be out of sync and cancel out nothing. Maybe some ML system could do this in the future, but it would have to be absolutely perfect to not sound distractingly strange at times.
Fans on the other hand are predictable, so you can output an opposing waveform despite the processing delay because you can predict how the waveform will act after your processing delay and keep them in phase.
Try noise cancellation on it. And results may be more clear. I did a project on active noise cancellation using neural networks. It should apply here as well.
That's the simple, signal-processing oriented way to noise cancellation.
But what if the software/ai is advanced enough to reproduce a sound, but erase a certain aspect of it? Like how photo editing can edit out an object or background? Then the headphones can use a seal to completely block out all noise, and play only the sounds the user selects!
Noise cancelling doesn't work with speech that well. It's way better for steady background noise such as on a small plane if you're the pilot or on a large plane if you're a passenger.
While the noise cancellation is active it will attempt to neutralize (destructively interfere with) sounds from the outside, including those generated by your speaker. You could indeed adversarially engage through something like a spontaneous phase shift (so the interference will become constructive, making the resulting signal louder) or generating a frequency the ANC can't compensate.
What you're describing is more or less why noise suppression algorithms in general cannot really improve intelligibility of the speech. Unless they're given extra cues (like with a microphone array), there's nothing they can do in real-time that will beat what the brain is capable of with "delayed decision" (sometimes you'll only understand a word 1-2 seconds after it's spoken). So the goal of noise suppression is really just making the speech less annoying when the SNR is high enough not to affect intelligibility.
That being said, I still have control over the tradeoffs the algorithm makes by changing the loss function, i.e. how different kinds of mistakes are penalized.
Active noise cancellation generally works best for consistent background noise. Sound that changes rapidly, like human voices, will be difficult for the cancellation hardware to react to, and headphones that provide sound isolation may be more help. See https://www.quora.com/Is-there-a-noise-cancelling-headphone-....
reply