Supervised training of object detectors requires well-annotated large-scale datasets, whose production is extremely expensive. Therefore, some efforts have been made to obtain annotations in economical ways such as cloud sourcing. However, datasets obtained by these methods tend to contain noisy annotations such as inaccurate bounding boxes and incorrect class labels. Our research thus focuses on training object detectors on datasets with entangled classification noise and localization annotation noise. In this study, we propose a framework to distinguish and correct the noisy annotations and subsequently train the detector using the corrected annotations. We verified the effectiveness of our proposed method and compared it with state-of-the-art methods on noisy datasets with different noise levels. The experimental results show that our proposed method significantly outperforms state-of-the-art methods.