Abstract: The rapid development of machine learning in recent years has provided new solutions for the back-end processing of optical fiber sensing data, which is of great significance for improving key indicators of sensors. The research progress of optical fiber sensors supported by machine learning and deep learning technologies was systematically reviewed. First, it focuses on the application of various machine learning algorithms and deep learning models in optical fiber sensing signal processing, feature extraction, and event recognition. Next, the strategies for converting 1-D signals into 2-D images were discussed, as well as the use of light spot image analysis methods to further enhance monitoring effectiveness. Finally, future research directions were outlined in areas such as multi-model fusion, hardware optimization, and edge computing, aimed at advancing the development of intelligent optical fiber sensing technology.