The youtubevis valid dataset

Epiphqny / VisTR

[CVPR2021 Oral] End-to-End Video Instance Segmentation with Transformers

https://arxiv.org/abs/2011.14503

Apache License 2.0

739 stars 95 forks source link

Closed sally1913105 closed 3 years ago

sally1913105 commented 3 years ago

I wanted to test the results of your model, but I found that there was no segmentation information in the label of YouTubevis's Valid dataset

sally1913105 commented 3 years ago

Did you combine the Valid label.json file with the Valid result.json file

Epiphqny commented 3 years ago

hi @sally1913105 , there is no ground truth for valid dataset, it is used as the test-dev to test the performance.

sally1913105 commented 3 years ago

Thank you for your answer. Do you split the train data as the valid dataset ? how can I to test your model to Verify the accuracy

Epiphqny commented 3 years ago

@sally1913105 As described in README, "by submitting the result json file to the CodeLab system"

sally1913105 commented 3 years ago

ok I get it