Effects of detection loophole on rival entanglement attestation techniques

Loopholes present in an experimental set-up can significantly affect the reliability of entanglement detection. We discuss two methods for detection of entanglement: one is by using the positive partial transposition criterion after quantum state tomography and the other by estimating the second and third moments of partial transposition of the quantum state through random classical snapshots. We examine the impact of inaccuracies in these detection methods by considering presence of spurious clicks or suppression of valid clicks in the detectors. By comparing the two methods, we observe that the condition based on partial transposition moments is more robust to missing counts than the positive partial transposition criteria. Moreover, we realize that in the presence of additional counts, none of the criteria misinterpret any separable state as entangled. But in such a scenario, the condition based on the moments can not guarantee any state as entangled, unless the additional event efficiency is about 0.9 or higher.