Dual-branch teacher-student with noise-tolerant learning for domain adaptive nighttime segmentation

Ruiying Chen, Yunan Liu, Yuming Bo, Mingyu Lu

Abstract: Highlights•We introduce a dual-branch teacher-student framework for unsupervised nighttime semantic segmentation.•The teacher and student networks collaborate to distill more reliable high-quality pseudo-labels.•We incorporate noise-tolerant learning to enhance the utilization of distilled pseudo-labels.•Our method achieves a new state-of-the-art performance when evaluated on benchmark datasets.