When Noise Meets Long-Tail: Feature-Threshold Dual Calibration for Robust Pseudo-Labeling
The results establish that jointly calibrating features and thresholds is essential for robust pseudo-labeling under compounded noise and class imbalance, and propose FTC-Seg, a Feature-Threshold dual-Calibration framework built on a standard teacher-student framework.
Ping-Song Guo, Zhi-Qi Huang, Xin-Ran Li
· 1 citation