search

LiheYoung / UniMatch

[CVPR 2023] Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic Segmentation
https://arxiv.org/abs/2208.09910
MIT License
478 stars 60 forks source link

use of unlabelled data #79

Closed kushiluv closed 1 year ago

kushiluv commented 1 year ago

Hey, i am a complete beginner but i'm trying to run your model in my own dataset which has a lot of unlabelled data. From my understanding the point of SSL algos is when you have a lot of unlabelled data which can also be used. But for the model to work I need to place the masks as well in unlabeled.txt (maybe it is being used to evaluate) , is there a way around it?

Liumomo30 commented 6 months ago

Hello, I have the same issue. Have you resolved it?

79zzz commented 6 months ago

你们指的是无标签数据的掩码吗？