RenYurui / Global-Flow-Local-Attention

The source code for paper "Deep Image Spatial Transformation for Person Image Generation"
https://renyurui.github.io/GFLA-web
Other
568 stars 84 forks source link

training and testing on own dataset #95

Open sparshgarg23 opened 2 years ago

sparshgarg23 commented 2 years ago

Hi,I was going through your repository and paper,and I would like to try it on my own dataset and task.

The task is to transform a source person image to rest pose image. In order to train and test on my dataset ,are these the following steps which I should follow 1.Split data into train,test and val.Is it necessary to create train.test and val.csv aftet the split.Also previous work on the deep fashion dataset required images to be resized.Is that applicable here also.Plus how to generate train.lst and test.lst? 2.Use OpenPose to obtain keypoint annotations 3.Generate pairs.csv 4.Finally run python script/generate_fashion_datasets.py Once this is done,I am assuming that I can then run train.py Apart from this is there anything else which I am missing