Llq201809

Thank you!👍------------------ 原始邮件 ------------------ 发件人: "Xiangshen M" 发送时间: 2019年5月17日(星期五) 晚上8:05 收件人: "kuaikuaikim/DFace"; 抄送: "Llq201809";"Author"; 主题: Re: [kuaikuaikim/DFace] TypeError: float() argument must be a stringor a number, not 'map' (#29) I...

> 我的当前环境ubuntu16; > > > > > import torch > > > > torch.**version** > > > > '1.0.1.post2' > > > > python -V > > > > Python...

链接：https://pan.baidu.com/s/13mvFLyKcnyIpF9dcNIvssw 提取码：jdat

请问，这个代码训练时使用了验证集，测试集标注不公开，所以我还是用的验证集去测试，但是效果不好。您们是怎么测试的widerface呢？ @1529046970 @yxlijun

you don't need to modify max_im_shrink if out of memory at /pytorch/torch/lib/THC/generic/THCStorage.cu:58, just change x = Variable(torch.from_numpy(x).unsqueeze(0)) to : x = Variable(torch.from_numpy(x).unsqueeze(0),volatile=True), it works.

> In the function of nms, there is an exception return needs one more value > > ``` > keep = scores.new(scores.size(0)).zero_().long() > if boxes.numel() == 0: > return keep...

Thanks for your reply very much. I find error in my data set. Now it works . @yangxue0827

I have writed a print function to delete error coordinate when translate dataset to VOC format. For example,if Xmin

Sorry，it should be Xmax-Xmin=0，delete it. ------------------ 原始邮件 ------------------ 发件人: "Rachquazar" 发送时间: 2019年8月30日(星期五) 下午5:57 收件人: "DetectionTeamUCAS/Cascade-RCNN_Tensorflow"; 抄送: "Llq201809";"Mention"; 主题: Re: [DetectionTeamUCAS/Cascade-RCNN_Tensorflow] rpn_loc_loss is nan,fast_rcnn_loc_loss is 0.0 (#8) @Llq201809 in my case,...