论文标题
日期：端到端完全卷积对象检测的双分配
DATE: Dual Assignment for End-to-End Fully Convolutional Object Detection
论文作者
论文摘要
完全卷积检测器丢弃一对一的任务，并采用一对一的分配策略来实现端到端检测，但遭受了缓慢的收敛问题。在本文中，我们重新审视了这两种分配方法，并发现将一对多分配带回端到端完全卷积探测器有助于模型收敛。基于此观察，我们提出{\ em \ textbf {d} ual \ textbf {a} ssignment}，用于端到端完全卷积de \ textbf {te} ction（date）。我们的方法在培训过程中构建了两个分支，并通过提供更多的监督信号来构建一个一对多和一对一的任务，并加快了一对一任务分支的融合。日期仅将分支与模型推理的一对一匹配策略一起使用，这不会带来推理开销。实验结果表明，双重分配给ONENET和DEFCN上提供了非平凡的改进，并加快了模型收敛的速度。代码：https：//github.com/yiqunchen1999/date。
Fully convolutional detectors discard the one-to-many assignment and adopt a one-to-one assigning strategy to achieve end-to-end detection but suffer from the slow convergence issue. In this paper, we revisit these two assignment methods and find that bringing one-to-many assignment back to end-to-end fully convolutional detectors helps with model convergence. Based on this observation, we propose {\em \textbf{D}ual \textbf{A}ssignment} for end-to-end fully convolutional de\textbf{TE}ction (DATE). Our method constructs two branches with one-to-many and one-to-one assignment during training and speeds up the convergence of the one-to-one assignment branch by providing more supervision signals. DATE only uses the branch with the one-to-one matching strategy for model inference, which doesn't bring inference overhead. Experimental results show that Dual Assignment gives nontrivial improvements and speeds up model convergence upon OneNet and DeFCN. Code: https://github.com/YiqunChen1999/date.
