In most of electronic and synth-pop music the accompaniments are played in very high or very low octaves that the voice couldn't match, so how the singers in this genres sing in tune without any accompaniment that suit their ranges ?
The singer doesn't need an explicit 'note to sing along to', it's done by relative pitch memory, practise & familiarity with the song.
This is not really all that different to the mechanism by which a classically-trained singer can sing from a sheet of dots. They just have a better set of internal references to rely on. For the rest of us, an opening chord is usually a good start.
"Reference pitch" doesn't need an octave. It doesn't even really need the exact note you're about to sing, if you know where it falls within the chord/key you're listening to. Take a simple example - Snow Patrol-Chasing Cars
[I found one with the entire intro to get you better settled. Starts at about 7 seconds in, set in the link.]
The intro is just 2 notes, A & E. The opening vocal line goes from quite a throwaway E below, C♯, G♯, A. By the time it actually hits A [our grounding "root" note], the accompaniment has moved to G♯ & E… so they never even meet. The singer is using a known structure to be able to know the notes not stated in the accompaniment.
The song in its entirety only uses two chords, and sits solidly in A major the whole way through. Starting in Amaj but with a passing major 7th in the bass [the G♯] before going to Dmaj, then back to the Amaj & then round again. It's all held together by the top E, which is a 5th in A but then a 9th in D, giving that 'haunting' repetition the whole piece depends on.
Once you know the base key & current chord, you 'know' where you need to be in relation to that. The more times you sing something, of course, the easier that becomes.