[CVPR'22 & IJCV'24] Semi-Supervised Semantic Segmentation Using Unreliable Pseudo-Labels & Using Unreliable Pseudo-Labels for Label-Efficient Semantic Segmentation
436
stars
61
forks
source link
方法同样适用于二分类任务吗? #157
Closed
ZHIZIHUABU closed 11 months ago
不太适用。二分类任务中定义 negative keys 比较困难