cf2

参考
https://blog.csdn.net/crazyice521/article/details/65935753?utm_medium=distribute.pc_relevant.none-task-blog-2%7Edefault%7EBlogCommendFromBaidu%7Edefault-6.control&dist_request_id=&depth_1-utm_source=distribute.pc_relevant.none-task-blog-2%7Edefault%7EBlogCommendFromBaidu%7Edefault-6.control

https://blog.csdn.net/susansmile1014/article/details/77644922?utm_medium=distribute.pc_relevant.none-task-blog-2_defaultBlogCommendFromMachineLearnPai2_{default-4.control&dist_request_id=&depth_1-utm_source=distribute.pc_relevant.none-task-blog-2}default_{BlogCommendFromMachineLearnPai2}default-4.control

https://blog.csdn.net/cakaf/article/details/108813843

https://blog.csdn.net/weixin_30552811/article/details/98752056?ops_request_misc=%257B%2522request%255Fid%2522%253A%2522161785088816780265421998%2522%252C%2522scm%2522%253A%252220140713.130102334.pc%255Fall.%2522%257D&request_id=161785088816780265421998&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2_allfirst_rank_v2~rank_v29-1-98752056.pc_search_result_no_baidu_js&utm_term=cf2%E4%BB%A3%E7%A0%81

1.下载代码
2.下载预训练模型imagenet-vgg-verydeep-19.mat(看/model目录下readme.txt)，到/model目录下
3.第一次跑要运行demo.m来编译matconvnet，之后只运行run_tracker.m就可

代码中有测试的视频序列在data里面，只是测试不需要下载OTB

4.报错

问：调用 “tracker_ensemble” 时，未对输出参数 “positions” (可能还包括其他参数)赋值。
解：把tracker_ensemble函数里的positions改为rects