Jianwei Yang

@sunilrao43 , Hello, you can extract the testing codes which do not rely on groundtruth landmarks inputs. Alternatively, you can make a fake pts file for your own images, so...

@Johnccl I will put the model on Baidu drive as well. The vt server seems not usable anymore.

Hi, there are roughly two steps: 1. register your dataset following the sample code in: https://github.com/UX-Decoder/Segment-Everything-Everywhere-All-At-Once/tree/v1.0/datasets/registration 2. create a dataset mapper following the sample code in: https://github.com/UX-Decoder/Segment-Everything-Everywhere-All-At-Once/tree/v1.0/datasets/dataset_mappers you can refer...

it's at the end of loadsamples.m

You do not have to get the groundtruth landmarks. If you have no groundtruth landmarks, you can remove Te_Data{k}.shape_gt from that line. 在 2014-12-08 11:02:01，"weiyt2014" [email protected] 写道： In globalpredicition.m,"drawshapes(Te_Data{k}.img_gray, [Te_Data{k}.shape_gt...

yes, if it is true, you may need to re-train the model. By th way, what is your training error?

Hi, thanks for your interest. So you are using the C++ project, right? Please give me some time to check it. Because i suppose the C++ codes is not aimed...

ok, I see, I will help to figure out why it did not work after changing the number of landmarks. On Wed, Sep 9, 2015 at 10:58 PM, sidc9 [email protected]...

I am very sorry that I am a little busy these days, I have not yet found the causes. Maybe you can also send email to Yuanxun Wang, his email...

@sidc9 I am very sorry that I am a little busy these days, I have not yet found the causes. Maybe you can also send email to Yuanxun Wang, his...