How to train and evaluate on deeplabv3 baseline?

Gaoyiminggithub / Graphonomy

Graphonomy: Universal Human Parsing via Graph Transfer Learning

MIT License

295 stars 68 forks source link

How to train and evaluate on deeplabv3 baseline? #22

Closed dragen1860 closed 3 years ago

dragen1860 commented 4 years ago

Dear author, Thanks for your neat implementation. I found the report deeplabv3 performance in this paper is relatively close to Graphonomy. Therefore I want to reproduce the deeplabv3 related experirments. Could you kindly give some tips on how to train and evaluate on deeplabv3 baselines using your repo.? Thank you.

Gaoyiminggithub commented 4 years ago

Hi, sorry for the late response. You can use the class in deeplab_xception.py to train the deeplabv3.

Gaoyiminggithub commented 3 years ago

Close due to long time no reply.