How can you imagine a photon? Is it correct to imagine a photon as a superposition of states <photon + virtual electron-positron pair + virtual quark + antiquark + ...> pair? And the lower the photon energy, the lower the amplitude of the virtual terms?
 A: In mainstream physics a photon is part of the elementary particle table which is axiomatically assumed in creating the quantum field theory standard model. The photon as all the particles is a point particle, of zero mass, and in QFT is created by creation operators acting on the photon field. The photon field, the electron field etc are quantum fields postulated over all spacetime, represented by the plane wave wave function of the corresponding differential equations: Dirac for fermions, Klein Gordon .... For photons it is the quantized Maxwell equation. .
This for interactin photons.A free photon cannot be described by a plane wave since a plane wave wavefunction  covers all spacetime. One uses the wave packet mathematics, so a free photon instead of having a unique frequency will have a probability distribution of many frequencies in the wave packet.
In no way a different model as you suggest, fits the observed data and behavior of a photon.
A: One should never think of a photon in terms of a "point-like particle", as it's neither point-like nor particle. First it is delocalised, any attempt to prescribe it a position is not Lorentz invariant, therefor since you cannot give a unique position in all reference frames we can't speak of point-like. This misuse of terminology comes from the fact that  measurement(the act of absorption) occurs at a specified point.
It is also not a particle, this interpretation streams from scattering experiments where the momentum transfer imitates particle scattering, from which one can conclude that the photon is as much of a particle as your strawberry ice cream is a strawberry fruit.
It is a quasi-partice, with all the mysticism this word brings, it can take different quantum states, can interfere, can tunnel, etc. Upon imagining a photon, one should define it properly, is it the object deduced from the equations with energy "

ℏ
ω

"
or the object that makes the detector clicks, as the two are not the same. I personally imagine the second since no source produces the first. The frequency cannot be defined by infinite precision, as in the real world no source has infinitely short lifetime, which brings us to the point that we always have a spread of frequencies(although small). This philosophical gymnastics makes me think of it as a collection of EM waves with a spread of 
Δω
, that can "sense" each other, which is the property of coherence. In my opinion this ontological interpretation of the photon is much closer to the object that Roy Glauber describes, consistent with Lamb Jr.'s criticism of the terminology and much closer to what the empirical reality suggests.
A: In simple imagination Photon is the packet of energy. The rest mass of Photon is zero, having energy E= hv(nu) , where is planks constant. Photon is moving with the velocity of light.
