论文标题
hit-detector：分层三位一体体系结构搜索对象检测
Hit-Detector: Hierarchical Trinity Architecture Search for Object Detection
论文作者
论文摘要
神经体系结构搜索（NAS）在图像分类任务中取得了巨大成功。最近的一些作品设法探索了有效的主链或特征融合层的自动设计，以进行对象检测。但是，这些方法的重点是仅搜索对象检测器的一个特定组成部分，同时手动设计其他组件。我们确定搜索组件和手动设计的组件之间的不一致性将保留更强性能的检测器。为此，我们提出了一个分层三位一体搜索框架，以同时发现对象检测器的所有组件（即骨架，颈部和头部）的有效体系结构，以端到端的方式。此外，我们从经验上揭示了检测器的不同部分更喜欢不同的操作员。在此激励的情况下，我们采用了一种新颖的方案来自动筛选不同组件的不同子搜索空间，以便有效地对相应的子搜索空间上的每个组件进行端到端搜索。如果没有铃铛和哨声，我们的搜索架构，即hit-detector，可以在具有27m参数的可可零售现场上获得41.4 \％的地图。我们的实施可在https://github.com/ggjy/hitdet.pytorch上获得。
Neural Architecture Search (NAS) has achieved great success in image classification task. Some recent works have managed to explore the automatic design of efficient backbone or feature fusion layer for object detection. However, these methods focus on searching only one certain component of object detector while leaving others manually designed. We identify the inconsistency between searched component and manually designed ones would withhold the detector of stronger performance. To this end, we propose a hierarchical trinity search framework to simultaneously discover efficient architectures for all components (i.e. backbone, neck, and head) of object detector in an end-to-end manner. In addition, we empirically reveal that different parts of the detector prefer different operators. Motivated by this, we employ a novel scheme to automatically screen different sub search spaces for different components so as to perform the end-to-end search for each component on the corresponding sub search space efficiently. Without bells and whistles, our searched architecture, namely Hit-Detector, achieves 41.4\% mAP on COCO minival set with 27M parameters. Our implementation is available at https://github.com/ggjy/HitDet.pytorch.
