数据处理

[fAnchao1]

Please Download the zip file if you do not want to generate json files yourself. It includes three json files: ilsvrc_vid.txt, youtube_final.txt, vot2018.txt. In our training process, we use ilsvrc_vid.txt and youtube_final.txt to train and vot2018.txt to verify the model.

Try python label_preprocess.py FILE OUTPUTFILE PATH to do the preprocess for your *.txt. For example, if you download the VOT2018 dataset in the folder /root/myfolder/ and the original file is vot2018.txt and the output file is vot2018_new.txt, then your command should be python label_preprocess.py vot2018.txt vot2018_new.txt /root/myfolder/

大佬，我想问下，这两个步骤是任选其一还是都得做

[leeyeehoo]

两部都做