论文翻译——YOLOv3: An Incremental Improvement

摘要

翻译：

我们向YOLO提供一些更新!我们做了一些设计上的小改变来让它更好。我们还训练了这个非常棒的新网络。它比上次大了一点，但更准确。不过它还是很快的，别担心。在320×320 的尺寸下YOLOv3运行了22 ms 在28.2 mAP时，与SSD一样准确，但三倍快。当我们看旧的0.5 IOUmAP检测度量YOLOv3是相当好的。它在Titan X上实现57.9 AP50在 51 ms，相比之下，RetinaNet的57.5 AP50in 198 ms，性能类似，但更快3.8倍。和往常一样，所有代码都可以在线访问https://pjreddie.com/yolo/。

 

原文: 可修改后右键重新翻译

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 320 × 320 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 AP50in 51 ms on a Titan X, compared to 57.5 AP50in 198 ms by RetinaNet, similar performance but 3.8× faster . As always, all the code is online at https://pjreddie.com/yolo/.

1 说明

翻译：

有时候你只是给它打电话，你知道吗?今年我没有做很多研究。花很多时间在推特上。和甘斯玩了一会儿。去年剩下了一点动力[12][1];我对YOLO做了一些改进。但是，说实话，没有什么特别有趣的，只是一些小的改变使它更好。我也帮助别人做一些研究。事实上，这就是我们今天来到这里的原因。我们有一个拍摄截止日期[4]，我们需要引用一些我随机更新给YOLO，但我们没有来源。所以准备好接受一份技术报告吧!科技报道的好处是他们不需要介绍，你们都知道我们为什么在这里。因此，这篇介绍的结尾将是本文其余部分的路标。首先我们告诉你YOLOv3的交易是什么。然后我们会告诉你我们是怎么做的。我们还会告诉你一些我们尝试过但没有成功的事情。最后，我们将思考这一切意味着什么。

 

Sometimes you just kinda phone it in for a year, you know? I didn’t do a whole lot of research this year. Spent a lot of time on Twitter. Played around with GANs a little. I had a little momentum left over from last year [12] [1]; I managed to make some improvements to YOLO. But, honestly, nothing like super interesting, just a bunch of small changes that make it better. I also helped out with other people’s research a little. Actually, that’s what brings us here today. We have a camera-ready deadline [4] and we need to cite some of the random updates I made to YOLO but we don’t have a source. So get ready for a TECH REPORT! The great thing about tech reports is that they don’t need intros, y’all know why we’re here. So the end of this introduction will signpost for the rest of the paper. First we’ll tell you what the deal is with YOLOv3. Then we’ll tell you how we do. We’ll also tell you about some things we tried that didn’t work. Finally we’ll contemplate what this all means.

2 The Deal

翻译：

YOLOv3的做法是:我们基本上都是借鉴别人的好点子。我们还训练了一个新的分类器网络，它比其他的分类器网络更好。我们会从头开始给你介绍整个系统，这样你就能完全理解了。

 

So here’s the deal with YOLOv3: We mostly took good ideas from other people. We also trained a new classifier network that’s better than the other ones. We’ll just take you through the whole system from scratch so you can understand it all.

2.1 边界框的预测

翻译：

在YOLO9000之后，我们的系统使用维度聚类作为锚盒[15]来预测边界盒。该网络为每个边界盒预测了4个坐标，tx, ty, tw, th。如果单元格与图像左上角的偏移量为(cx, cy)，且边界盒先验的宽度和高度为pw, ph，则预测值为:

Following YOLO9000 our system predicts bounding boxes using dimension clusters as anchor boxes [15]. The network predicts 4 coordinates for each bounding box, tx, ty, tw, th. If the cell is offset from the top left corner of the image by (cx, cy) and the bounding box prior has width and height pw, ph, then the predictions correspond to:

翻译：

在训练中，我们使用平方和误差损失。如果我们ground truth对坐标的预测是 ˆt*，那么我们的梯度是ground truth（从ground truth框计算)减去我们的预测：ˆt*− t*。通过反求上述方程，可以很容易地计算出该gr