Direct imaging is rapidly emerging as a promising technique to discover exoplanets. Contrary to transits and radial velocities, direct imaging is preferentially sensitive to young giant planets in wide orbits. This little-studied class of exoplanets looks especially interesting, since it can shed light on the still poorly-constrained early phases of planet formation, with important consequences on the architecture and the demographics of more evolved planetary systems. Being detected in the NIR through their thermal emission -related to their residual formation heat-, these planets appear as brighter sources if younger; but, of course, more massive planets are brighter too. To break the age-mass degeneracy, constraining the age of the stellar host is of utmost importance. Here we show how, starting from astrometric data from Gaia EDR3 and radial velocities from GALAH and APOGEE, we identified subgroups of coeval stars within the 5-11 Myr old Upper Scorpius association employing a novel kinematic method (Squicciarini et al. 2021). These results are particularly valuable for the ongoing planet-hunting BEAST survey (Janson et al. 2021), as strong constraints on the ages of the stellar targets are needed to correctly interpretate the forthcoming planet detections.