论文标题
教学：与老师的更好培训
Teach-DETR: Better Training DETR with Teachers
论文作者
论文摘要
在本文中，我们介绍了一种新颖的培训计划，即教学范围，以从多功能教师探测器中学习更好的基于DITR的探测器。我们表明，来自教师探测器的预测框是传递教师探测器知识的有效培养基，这些探测器可以是基于RCNN或基于DETR的探测器，以训练更准确，更强大的DETR模型。这种新的培训计划可以轻松地合并来自多个教师探测器的预测框，每个探测器都为学生培训提供了平行的监督。我们的策略没有引入其他参数，并且在培训期间为原始检测器增加了可忽略的计算成本。在推断期间，Teach-detr会带来零额外的开销，并保持不需要非最大抑制作用的优点。广泛的实验表明，我们的方法会导致各种基于DITR的检测器的一致改进。具体而言，我们使用Swin-Large主链，4个特征地图和36个上类培训时间表的最先进的探测器Dino从57.8％提高到58.9％，而MSCOCO 2017验证集的平均平均精度。代码将在https://github.com/leonhlj/teach-detr上找到。
In this paper, we present a novel training scheme, namely Teach-DETR, to learn better DETR-based detectors from versatile teacher detectors. We show that the predicted boxes from teacher detectors are effective medium to transfer knowledge of teacher detectors, which could be either RCNN-based or DETR-based detectors, to train a more accurate and robust DETR model. This new training scheme can easily incorporate the predicted boxes from multiple teacher detectors, each of which provides parallel supervisions to the student DETR. Our strategy introduces no additional parameters and adds negligible computational cost to the original detector during training. During inference, Teach-DETR brings zero additional overhead and maintains the merit of requiring no non-maximum suppression. Extensive experiments show that our method leads to consistent improvement for various DETR-based detectors. Specifically, we improve the state-of-the-art detector DINO with Swin-Large backbone, 4 scales of feature maps and 36-epoch training schedule, from 57.8% to 58.9% in terms of mean average precision on MSCOCO 2017 validation set. Code will be available at https://github.com/LeonHLJ/Teach-DETR.
