The rise of Fintech has produced a variety of innovative financial services. In this work, we present a system for mobile phone financial services, which can be utilized by users registering their own accounts and for accessing the created account using their own faces. In the proposed system, face recognition and landmark detection methods are used to achieve identity recognition and liveness detection tasks from face images captured by a mobile phone. However, sometimes malicious actors attempt to use a face photo of owners to illegally access their accounts. To prevent the illegal access, we further design a mechanism for users to interact with the system and access their accounts followed by random instructions. In the experiment, the accuracy and execution time results were found to be acceptable. In addition, we further recruited subjects for the purpose of a subjective evaluation using the System Usability Scale (SUS) and the Technology Acceptance Model (TAM), and the results showed the proposed system is acceptable for use.