Put $n = 2014 = 2.19.53$. By the Sylow theorems, the number $n_p$ of Sylow $p$-subgroups of $G$ has $n_p|(n/p)$ and $n_p \equiv 1\pmod{p}$ for $p = 2, 19, 53$. It follows that $n_p = 1$ for $p = 19, 53$; let $C_p \simeq \mathbb{Z}/p\mathbb{Z}$ denote the Sylow $p$-subgroup of $G$. Clearly $C_{19}\cap C_{51} = 1$, so the group $K = C_{19} C_{51} \subset G$ is isomorphic to $\mathbb{Z}_{19} \times \mathbb{Z}_{51}$. Furthermore, the groups $C_{19}, C_{51}$ are normal in $G$ by the Sylow theorems, so $K$ is as well. For any $g\in G$ of order $2$ (which must exist, by the Sylow theorems), we have $\langle{g, K\rangle} = G$ with $g$ acting on $K$ by conjugation. Thus $G$ is a semidirect product $\langle{g\rangle}\ltimes K = \mathbb{Z}_2 \ltimes (\mathbb{Z}_{19}\times \mathbb{Z}_{51})$ with respect to some action $\mathbb{Z}_2 \to \operatorname{Aut}(\mathbb{Z}_{19}\times \mathbb{Z}_{51})$. (If you're not familiar with semidirect products, I'd recommend you look up the construction; they commonly pop up in problems like this, and it's easier to work from the definition rather than trying to describe it ad hoc.)
So, now we need to compute $\operatorname{Aut}(\mathbb{Z}_{19}\times \mathbb{Z}_{51}) = \operatorname{Aut}(\mathbb{Z}_{n/2})$. The automorphism group of a cyclic group $\mathbb{Z}_m$ is clearly simply $\mathbb{Z}_m^{\times}$, comprising the maps $g \to g^k$ with $(k, m) = 1$ for some fixed generator $g\in \mathbb{Z}_m$. The group of maps $\mathbb{Z}_2 \to \operatorname{Aut}(\mathbb{Z}_{19} \times \mathbb{Z}_{51})$ thus comprises the four maps sending a generator to $(g, h) \to (g^{\alpha}, h^{\beta})$, where $\alpha^2 = 1$ in $\mathbb{Z}_{19}$ and $\beta^2 = 1$ in $\mathbb{Z}_{51}$. This construction gives the required four groups; to show that they're distinct, consider the centralizer of an element of order $2$.
The same approach works in general, though it depends on having $n$ be the product of convenient primes. The technique is honestly not that powerful, but this kind of computation appears on a test in every single class I've seen covering the Sylow theorems, and even some candidacy exams. In this particular case, you're lucky that the divisors of $n$ are reasonably spaced primes, so that the combinatorial constraints imposed by the Sylow theorems are so limiting; this sort of problem often involves some extra complications with solvability, prime powers in the order of $G$, etc.