Person re-identification is a hot topic in the field of computer vision. In the long-term re-identification, due to the change of person clothes, person re-identification has become a difficult task. In view of the above difficulty, this paper proposes a dual-category recognition method for clothes-changing people, and use spatial attention mechanism to capture important information in the image. Then, the sensitivity awareness of clothes changes is further proposed to expand the clothes differences between categories and reduce the interference of similar clothes in clothes classification. This re-identification method based on clothes sensitivity awareness and spatial attention mechanism can realize the comprehensive extraction of clothes-independent features. Extensive experiments on benchmark datasets have confirmed the effectiveness of the method.