YOLOv3: An Incremental Improvement

原文地址:YOLOv3: An Incremental Improvement

Abstract

We present some updates to YOLO! We made a bunch of little design changes to make it better. We also trained this new network that's pretty swell. It's a little bigger than last time but more accurate. It's still fast though, don't worry. At 320x320 YOLOv3 runs in 22 ms at 28.2 mAP, as accurate as SSD but three times faster. When we look at the old .5 IOU mAP detection metric YOLOv3 is quite good. It achieves 57.9 mAP@50 in 51 ms on a Titan X, compared to 57.5 mAP@50 in 198 ms by RetinaNet, similar performance but 3.8x faster. As always, all the code is online at https://pjreddie.com/yolo/.

我们对YOLO进行了一些更新!通过一系列小的设计更改可以使其变得更好。我们训练了这个非常棒的新网络,它比上次稍大,但更准确。不过还是很快,别担心。输入320x320图像,YOLOv3运行时间为22ms, 精度为28.2 mAP,和SSD一样准确,但速度快三倍。使用旧的0.5 IOU mAP检测指标时,YOLOv3可以在Titan X上达到57.9 mAP@50,单次推理时间为51 ms,相比之下RetinaNet需要198ms,精度为57.5 mAP@50,速度快了3.8倍。跟以往一样,所有代码都开源在https://pjreddie.com/yolo/.

解析

从接