In the last chapter, we introduced research on 3D facial aesthetics. As an effective alternative, 3D facial reconstruction techniques greatly reduce the difficulty of acquiring 3D aesthetic datasets. Besides, 3D facial reconstruction can be used in multimedia applications such as facial editing and digital avatars which requires not only geometric accuracy but also aesthetic consistency. In this chapter, we will introduce the principle of 3D reconstruction and some classical reconstruction approach. However, existing methods prioritize accuracy while neglecting aesthetics, which can diminish user experience in visually demanding tasks. Thus, we will show our previous study, which introduces facial aesthetics into the 3D reconstruction process, using a hybrid loss semi-supervised approach that combines identity and beauty supervision. We analyzed how the proposed hybrid loss impacts the reconstruction results theoretically and introduced a beauty consistency metric to measure reconstruction aesthetic consistency. By integrating hybrid loss supervision, our method enhances both reconstruction accuracy and facial aesthetic consistency. We compare the state-of-the-art methods and the experimental results demonstrate superior performance of proposed model on NoW, REALY benchmarks and beauty consistency metric, highlighting the importance of aesthetics in 3D face reconstruction.