Triply periodic minimal surfaces (TPMS) metamaterials and shape-memory polymer (SMP) smart materials are known for their beneficial attributes in novel scientific and industrial fields. Through TPMS designs, low weight accompanied by high surface area are achievable, which are known as crucial parameters in many fields, such as tissue engineering. Moreover, SMPs are well-suited to generate force or to recover their permanent shape by means of an external stimulus. Combining these properties is possible by fabricating TPMS-based metamaterials made out of SMPs, which can be applicable in numerous applications. By considering different level volume fraction of four types of TPMS-based lattices (diamond, gyroid, IWP, and primitive), we focus on the effect of micro-architecture on shape-memory characteristics (i.e., shape recovery, shape fixity, and force recovery) as well as mechanical properties (elastic modulus and Poisson's ratio) of these smart metamaterials. For this purpose, shape-memory effect (SME) is simulated employing thermo-visco-hyperelastic constitutive equations coupled with the time-temperature superposition principle. It is observed that by increasing the level volume fraction of each lattice type, the elastic modulus, shape fixity, and force recovery increase, while the shape recovery diminishes. Such behaviors can be attributed to different deformation modes (flexural or uniaxial) in SMP TPMS-based metamaterials. Furthermore, it is shown that the Poisson's ratio has a nonlinear behavior in these structures. The smart metamaterials introduced in this study have the advantage of providing the possibility of designing implants, especially in bone defects tailored with different micro-architectures depending on each patient's specific need.