Day-to-day observations reveal numerous medical and social situations where maintaining physical distancing is either not feasible or not practiced during the time of a viral pandemic, such as, the coronavirus disease 2019 (COVID-19). During these close-up, face-to-face interactions, a common belief is that a susceptible person wearing a face mask is safe, at least to a large extent, from foreign airborne sneeze and cough droplets. This study, for the first time, quantitatively verifies this notion. Droplet flow visualization experiments of a simulated face-to-face interaction with a mask in place were conducted using the particle image velocimetry setup. Five masks were tested in a snug-fit configuration (i.e., with no leakage around the edges): N-95, surgical, cloth PM 2.5, cloth, and wetted cloth PM 2.5. Except for the N-95 mask, the findings showed leakage of airborne droplets through all the face masks in both the configurations of (1) a susceptible person wearing a mask for protection and (2) a virus carrier wearing a mask to prevent the spreading of the virus. When the leakage percentages of these airborne droplets were expressed in terms of the number of virus particles, it was found that masks would not offer complete protection to a susceptible person from a viral infection in close (e.g., <6 ft) face-to-face or frontal human interactions. Therefore, consideration must be given to minimize or avoid such interactions, if possible. This study lends quantitative support to the social distancing and mask-wearing guidelines proposed by the medical research community.