search

bearpaw / PyraNet

Code for "Learning Feature Pyramids for Human Pose Estimation" (ICCV 2017)
Apache License 2.0
221 stars 58 forks source link

Evaluation on LSP Dataset #11

Closed MissYOLY closed 6 years ago

MissYOLY commented 6 years ago

Hi,@bearpaw Thank you for sharing your wonderful work! Based on the steps you provided, after training on the LSP data set, generate a .h5 prediction file, but in this evaluation code: http://human-pose.mpi-inf.mpg.de/results/lsp/evalLSP. Zip, it needs to use .mat files to make predictions. So what I want to ask is how to generate such a .mat file ? Have you implemented such functions in your codes?

738654805 commented 6 years ago

@MissYOLY how to train on LSP dataset.In this work,Options only include mpii and mpii-lsp.

MissYOLY commented 6 years ago

@738654805 I just set mpii-lsp for train LSP.

738654805 commented 6 years ago

@MissYOLY thanks for your reply.You mean the option mpii-lsp is equal to load lsp dataset?not mpii+lsp.

MissYOLY commented 6 years ago

The option mpii-lsp is equivalent to loading the lsp+mpii data set. Many methods are combined training with mpii and lsp, and then tested with the lsp test set.

MissYOLY commented 6 years ago

@738654805

738654805 commented 6 years ago

got it.thanks. BTW,do you have access to download the MPII+LSP data set.

MissYOLY commented 6 years ago

Yes, just change the option.

bearpaw commented 6 years ago

@MissYOLY Matlab can read h5 files. You can also use python to convert h5 to mat files. Just google it :)