Label assignment plays a significant role in modern object detection models. Detection models may yield totally different performances with different label assignment strategies. For anchor-based detection models, the IoU threshold between the anchors and their corresponding ground truth bounding boxes is the key element since the positive samples and negative samples are divided by the IoU threshold. Early object detectors simply utilize a fixed threshold for all training samples, while recent detection algorithms focus on adaptive thresholds based on the distribution of the IoUs to the ground truth boxes. In this paper, we introduce a simple and effective approach to perform label assignment dynamically based on the training status with predictions. By introducing the predictions in label assignment, more high-quality samples with higher IoUs to the ground truth objects are selected as the positive samples, which could reduce the discrepancy between the classification scores and the IoU scores, and generate more high-quality boundary boxes. Our approach shows improvements in the performance of the detection models with the adaptive label assignment algorithm and lower bounding box losses for those positive samples, indicating more samples with higher quality predicted boxes are selected as positives. The source code will be available at https://github.com/ZTX-100/DLA-Combined-IoUs.
翻译:在现代物体探测模型中, Label 任务在现代物体探测模型中起着重要作用。 检测模型可以产生完全不同的性能,使用不同的标签分配策略。 对于基于锚的检测模型,锚及其相应的地面事实约束框之间的IOU阈值是关键要素,因为正样和负样由IoU阈值除以。 早期物体探测器只是对所有培训样本使用固定阈值,而最近的检测算法则侧重于基于将IOU分布到地面真相框的适应性阈值。 在本文中,我们采用一种简单而有效的方法,根据预测的培训状态动态地执行标签分配。 通过在标签分配中引入预测,将更高IOU值的高质量样本选为正样,从而可以减少分类分数与IOU分数之间的差异,并产生更高质量的边界框。 我们的方法显示,根据适应性标签分配算法,检测模型的性能有所改进,并降低这些阳性标的框损失。 通过在标签分配中引入了更多质量预测框的样本,将具有更高的IUUUU-M-I/Ix100/Uxx/ comm 的源码选为正。