“The application of 3D point cloud data processing technology in fields such as autonomous driving, robotics, and high-precision maps is gradually becoming prominent. However, current processing methods mainly rely on large-scale high-quality annotated datasets, and the model's generalization performance is limited, which has become a major challenge in this field. To address this challenge, the academic community has begun exploring the application of domain adaptive learning in point cloud data processing. Domain adaptive learning, as an important branch of transfer learning, aims to improve the adaptability of models across different domains. Therefore, this article systematically reviews the adaptive learning methods in the 3D point cloud domain in recent years, mainly including adversarial learning, cross modal learning, pseudo label learning, and data alignment. Each method has its unique advantages and challenges, which provide important references for subsequent research. Overall, this study not only contributes to a deeper understanding of the adaptive learning field in the point cloud domain, but also provides new ideas for addressing the annotation dataset requirements and model generalization problems in 3D point cloud data processing. In the future, with the continuous advancement of technology, adaptive learning in the 3D point cloud domain is expected to play an important role in more fields.”