Object Detection (OD) is an important task in Computer Vision with many\npractical applications. For some use cases, OD must be done on videos, where\nthe object of interest has a periodic motion. In this paper, we formalize the\nproblem of periodic OD, which consists in improving the performance of an OD\nmodel in the specific case where the object of interest is repeating similar\nspatio-temporal trajectories with respect to the video frames. The proposed\napproach is based on training a Gaussian Process to model the periodic motion,\nand use it to filter out the erroneous predictions of the OD model. By\nsimulating various OD models and periodic trajectories, we demonstrate that\nthis filtering approach, which is entirely data-driven, improves the detection\nperformance by a large margin.\n