模型笔记

1、减小过拟合：数据增广、正则化、droupout、提前终止、bagging方法。https://mp.csdn.net/postedit/100112018

2、损失函数：logistics loss、softmax loss、cross entropy loss的比较。

https://blog.csdn.net/yuechuen/article/details/70953796 （公式有几处错误，见笔记）

https://blog.csdn.net/ziyouyi111/article/details/80598469

3、yolov1-v3模型https://www.cnblogs.com/makefile/p/YOLOv3.html

yolov2： https://blog.csdn.net/u014380165/article/details/77961414

yolov3： https://blog.csdn.net/leviopku/article/details/82660381

               https://blog.csdn.net/qq_37541097/article/details/81214953

               https://www.jianshu.com/p/d13ae1055302

4、SSD

https://blog.csdn.net/baidu_41848695/article/details/100023053

https://blog.csdn.net/u014696921/article/details/53698300

5、deep sort

特征提取网络论文笔记https://blog.csdn.net/TYUT_xiaoming/article/details/99773097

6、Inception

https://blog.csdn.net/u011021773/article/details/80791650

7、Faster RCNN 损失函数

https://blog.csdn.net/Mr_health/article/details/84970776