Coronal mass ejections (CMEs) can cause severe geomagnetic storms on Earth, making it essential to forecast their arrival. However, predicting the transit time (TT) of Earth-directed CMEs is challenging due to the complex factors involved, such as CME dynamics, the state of the interplanetary medium, and CME interactions. Our goal is to predict the TT of Earth-directed CMEs without relying on computationally expensive methods. CMEs observed from L1 are significantly affected by projection effects, which underestimate their speed and overestimate their size and mass. To address this, we use the Solar TErrestrial RElations Observatory (STEREO) to study the kinematics of halo and partial halo CMEs in quadrature and near-quadrature configurations. We analyzed a data set of 44 Earth-directed CMEs, dividing it into two categories: 25 CMEs in STEREO's quadrature field of view (FOV) and 19 CMEs in the near-quadrature FOV. Three different proxies were used to prepare empirical models for estimating TT. Our findings indicate that, for the considered sample of events, the empirical model based on the maximum speed of CMEs observed in STEREO's FOV yields the most accurate TT estimates. Our model achieved a maximum error of 16.46 hr and a mean error of 7.25 hr in estimating the TT of 44 Earth-directed CMEs.