In this paper, we propose an Android-based assistance system called SmileAtMe which uses smile detection as a means of direct interaction in order to rate meme images. A smartphone's front camera captures the user's face while they are interacting with the application on the device. A system evaluation uncovered that users fully understand how the application uses smile detection to rate images. The users attested that the automatic reaction classification works well. However, several of them were uncomfortable with the idea of being observed by their smartphone.