HPL-ESS: Hybrid Pseudo-Labeling for Unsupervised Event-based Semantic Segmentation
This work addresses the challenge of annotating event data for semantic segmentation, which is crucial for applications in high-speed motion and extreme lighting conditions, representing an incremental advancement in unsupervised domain adaptation.
The paper tackles the problem of noisy pseudo labels in unsupervised event-based semantic segmentation by proposing a hybrid pseudo-labeling framework, achieving a +5.88% accuracy and +10.32% mIoU improvement over state-of-the-art methods on the DSEC-Semantic dataset.
Event-based semantic segmentation has gained popularity due to its capability to deal with scenarios under high-speed motion and extreme lighting conditions, which cannot be addressed by conventional RGB cameras. Since it is hard to annotate event data, previous approaches rely on event-to-image reconstruction to obtain pseudo labels for training. However, this will inevitably introduce noise, and learning from noisy pseudo labels, especially when generated from a single source, may reinforce the errors. This drawback is also called confirmation bias in pseudo-labeling. In this paper, we propose a novel hybrid pseudo-labeling framework for unsupervised event-based semantic segmentation, HPL-ESS, to alleviate the influence of noisy pseudo labels. In particular, we first employ a plain unsupervised domain adaptation framework as our baseline, which can generate a set of pseudo labels through self-training. Then, we incorporate offline event-to-image reconstruction into the framework, and obtain another set of pseudo labels by predicting segmentation maps on the reconstructed images. A noisy label learning strategy is designed to mix the two sets of pseudo labels and enhance the quality. Moreover, we propose a soft prototypical alignment module to further improve the consistency of target domain features. Extensive experiments show that our proposed method outperforms existing state-of-the-art methods by a large margin on the DSEC-Semantic dataset (+5.88% accuracy, +10.32% mIoU), which even surpasses several supervised methods.