I have a lecture that was recorded poorly and from the back of the room, so picks up everyones movements close to it, but the lecturer you can hardly hear. Is there a way I can bring the lecturers voice to the foreground and make the voices and noises close to the microphone less audible..? Thank you. I've played around with noise reduction and then try to amplify the audio but it still comes out barely audible.
Short, honest answer: nope, forget it.
In principle it's possible nowadays to separate any part of a signal, using software like iZotope RX or the (rather dated, by now) Algorithmix Renovator. These are expensive heavyweight products; unfortunately I'm not aware of any more affordable or even open-source alternatives that get close to their performance.
And even in these programs, you'd still need a heck lot of manual effort to decide what to separate. The problem is that unlike white noise, other voices close to the microphone are hard to distinguish by software from the voice you want extracted. As yet, the human ear is by far the best at this task, so frankly, for the purpose of making the speaker better intelligible it's probably vain to use any software – better use a human with good headphones and a typewriter...
The latest version of RX actually has “neural network”, artificial intelligence features built in that are supposed to be able to accomplish some of these traditional human tasks with unprecedented precision. I haven't used any of this so I don't know how well it works, but I'd be surprised if it does work at all if you have yourself trouble even hearing the lecturer.
So, lesson learned: to get a good final recording, it's important that you start out with good source material! A lecture should always be recorded with a close mic (e.g. a lavalier) or at least with a shotgun mic pointed at the speaker.