论文标题
技术报告：一阶段轻量化对象探测器
Tech Report: One-stage Lightweight Object Detectors
论文作者
论文摘要
这项工作用于设计一阶段的轻型探测器，在地图和延迟方面表现良好。使用基线模型分别在GPU和CPU上进行靶向，因此在基线模型的骨干网络中应用了各种操作，而不是主要操作。除了有关骨干网络和操作的实验外，还研究了几种特征金字塔网络（FPN）体系结构。 MS COCO数据集上的参数，GFLOPS，GPU延迟，CPU潜伏期和MAP的数量分析了基准测试和提议的检测器，该参数，GPU，GPU延迟，CPU延迟和MAP在MS COCO数据集上是对象检测中的基准数据集。考虑到准确性和延迟之间的权衡，这项工作提出了类似或更好的网络体系结构。例如，我们提出的GPU-TARGET骨干网络的表现优于Yolox微型，该网络的速度为1.43倍，在NVIDIA GEFORCE RTX 2080 TI GPU上被选为1.43倍和0.5 MAP。
This work is for designing one-stage lightweight detectors which perform well in terms of mAP and latency. With baseline models each of which targets on GPU and CPU respectively, various operations are applied instead of the main operations in backbone networks of baseline models. In addition to experiments about backbone networks and operations, several feature pyramid network (FPN) architectures are investigated. Benchmarks and proposed detectors are analyzed in terms of the number of parameters, Gflops, GPU latency, CPU latency and mAP, on MS COCO dataset which is a benchmark dataset in object detection. This work propose similar or better network architectures considering the trade-off between accuracy and latency. For example, our proposed GPU-target backbone network outperforms that of YOLOX-tiny which is selected as the benchmark by 1.43x in speed and 0.5 mAP in accuracy on NVIDIA GeForce RTX 2080 Ti GPU.
