MovingObjectSegmentation
MovingObjectSegmentation copied to clipboard
Test split
I'm looking for the Test folder as specified regular_training.m
:
imgDir = ['../Test' video '/input'];
There is no reference in the code of the creation of the folder. How is the split made? The dataset file does not have this split.
Thanks
The test folder are the testing data from the CDNet2014 dataset
I have downloaded the CDNet2014 dataset here but there is no specific test folder. Could you point me to the file you have used?
All the frames for each video
获取 Outlook for iOShttps://aka.ms/o0ukef
收件人: Willams de Lima [email protected] 发送时间: 星期三, 十一月 14, 2018 11:53 下午 收件人: zhimingluo/MovingObjectSegmentation 抄送: Zhiming Luo; Comment 主题: Re: [zhimingluo/MovingObjectSegmentation] Test split (#4)
I have downloaded the CDNet2014 dataset here http://jacarini.dinf.usherbrooke.ca/dataset2014/ but there is no specific test folder. Could you point me to the file you have used?
― You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/zhimingluo/MovingObjectSegmentation/issues/4#issuecomment-438712164, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ALFRQwUDN5HvJ4HKFJ5r0YIUYNkeNZpyks5uvDyMgaJpZM4Yd6rs.
Sorry for the confusion, but you trained on all frames and tested on the same frames?
We selected around 100-200 frames for training, rest for testing, the details are in the paper
获取 Outlook for iOShttps://aka.ms/o0ukef
收件人: Willams de Lima [email protected] 发送时间: 星期四, 十一月 15, 2018 12:02 上午 收件人: zhimingluo/MovingObjectSegmentation 抄送: Zhiming Luo; Comment 主题: Re: [zhimingluo/MovingObjectSegmentation] Test split (#4)
Sorry for the confusion, but you trained on all frames and tested on the same frames?
― You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/zhimingluo/MovingObjectSegmentation/issues/4#issuecomment-438715639, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ALFRQ9Kvmw3eyRo1-gZoopJ-GMfC7AZKks5uvD6JgaJpZM4Yd6rs.