bbidong

still still waiting...

@Vickeyhw Hi , i meet the same error. I delete the cub and manual add it again, then the problem is solved.

yeap, I do the same with you. But I meet some problems during the process “make -j4”,------------------ 原始邮件 ------------------ 发件人: "lostork" 发送时间: 2019年10月5日(星期六) 晚上9:35 收件人: "msracver/Flow-Guided-Feature-Aggregation"; 抄送: "bbidong";"Mention"; 主题: Re:...

hi, i am sorry to be late， you can follow the lostork to solve the problem. ------------------ 原始邮件 ------------------ 发件人: "Steff Lin" 发送时间: 2019年9月25日(星期三) 下午5:26 收件人: "msracver/Flow-Guided-Feature-Aggregation"; 抄送: "bbidong";"Mention"; 主题:...

@alejodosr Hi, how do you set the batch size and lr ? And my loss is huge as follows. Are you same as me ? 

the evaluate result is lower than that in paper , because this is not the official code , so i try another vod method.------------------ 原始邮件 ------------------ 发件人: "KeepGoing2019HaHa"

I have not use this tensorflow code , maybe you can try "MEGA for Video Object Detection" which is a PyTorch implementation of a 2020 paper.