Liqian Ma

Hi, it seems the input data are not correct. Did you generate the test results successfully? On Sat, 31 Oct 2020 at 19:16, Emna B wrote: > hello, > when...

Hi, at that moment, I used the old version of OpenPose whose output format is not exactly the keypoint coordinates. For easier usage, I recommend checking run_convert_RCV.sh to convert the...

Hi Yao1up, these directories are the parent directories where you store the models, please check the following lines. https://github.com/charliememory/Disentangled-Person-Image-Generation/blob/e4703860bb1b351050ce50f339985ff0811f1d64/run_market_train.sh#L19 https://github.com/charliememory/Disentangled-Person-Image-Generation/blob/e4703860bb1b351050ce50f339985ff0811f1d64/run_market_train.sh#L48

I use the following scripts to switch to another gcc version and install the correlation lib successfully on CUDA 9.2 (original gcc-v8 not work) ``` # used to compile .cu...

@moha23 You can install it without sudo. While you need to make and compile from the gcc source code.

The bounding_box_test_clean is the name that comes from the original Market1501 dataset. I think it means to exclude the junk detections. The link is fixed now. On Tue, 1 Dec...

This issue comes from pickle package. To avoid this error, you can use "wb" mode and "rb" mode to dump and load data. On Fri, 6 Nov 2020 at 10:37,...

I think this may be due to the incompatibility between pickle-python2 and pickle-python3. The code and pickle data are developed under python2. Do you use python2 to load the data?...

There is a path error in your log "ValueError: The passed save_path is not a valid checkpoint: ./data/model/model.ckpt-0". Did you set the correct path? As to the second error, it...

Yes, the script is used for the uploaded pretrained models. You need to change the names if you test your own models. On Mon, 16 Nov 2020 at 09:38, Yao1up...