Interestingly, the dating of the Deir el medineh 'spangenhelm' to the 3rd-4th century seems to rely in part on the similarity to these helmets shown on the Arch of Galerius - it's also been dated (based on the hinges, I think) to the 6th century... So we should be cautious, perhaps. The helmets on the arch are also rather similar to those on the Adamklisi metopes, which we 'know' to be Imperial Gallic/Italic types (but which could also be some sort of spangenhelm...).
Actually, I've often thought that these helmets rather resemble
this one (particularly the bulbous top), but that's an Avar model from the 6th century...
The helmet on the left in the second picture looks a bit like those on the Aurelius Cervianus phalera - same brim or flange - which dates to the mid 3rd century. As with so many helmet types shown in later Roman art, I'm inclined to think that this might be something currently unknown in archaeology. Calling it an intercisa type might be convenient (after all, the intercisa looks fairly basic!), and we don't have much more to go on at present, but it's reductive even so.
The central helmet in the second picture could be something else again - I'm thinking of the Neiderbeiber type with an intercica style 'fin' crest found at Poitiers (as yet unpublished, I think) dating to c.270 - this is less than 30 years before the Persian campaign of Galerius, so these kinds of helmets could have still been in use.
Our knowledge of helmet types from this era is so patchy, and the conventions of Roman representational art so obscure, that we probably can't be any more exact about what we're seeing here.
EDIT - it may be useful to consider the identity of the troops on the first image. Might their distinctive helmets have been intended to represent a particular (e.g. Sarmatian?) origin?
Mailed Soldiers on the Arch of Galerius
(I think the suggestion that a mixed collection of soldiers from different units is intended is most persuasive, myself. But you never know...)