Class Balanced Dynamic Acquisition for Domain Adaptive Semantic Segmentation using Active Learning
Keywords: Active Learning, Class Balancing, Semantic Segmentation, Domain Adaptation
TL;DR: We introduce Class Balanced Dynamic Acquisition (CBDA), a new active learning approach for domain adaptive semantic segmentation that mitigates class imbalance, outperforming previous methods across different budgets.
Abstract: Domain adaptive active learning is leading the charge in label-efficient training of neural networks.
For semantic segmentation, state-of-the-art models jointly use two criteria of uncertainty and diversity to select training labels, combined with a pixel-wise acquisition strategy.
However, we show that such methods currently suffer from a class imbalance issue which degrades their performance for larger active learning budgets.
We then introduce Class Balanced Dynamic Acquisition (CBDA), a novel active learning method that mitigates this issue, especially in high-budget regimes.
The more balanced labels increase minority class performance, which in turn allows the model to outperform the previous baseline by 0.6, 1.7, and 2.4 mIoU for budgets of 5%, 10%, and 20%, respectively.
Additionally, the focus on minority classes leads to improvements of the minimum class performance of 0.5, 2.9, and 4.6 IoU respectively.
The top-performing model even exceeds the fully supervised baseline, showing that a more balanced label than the entire ground truth can be beneficial.
Submission Number: 12
Loading