HPL-ESS: Hybrid Pseudo-Labeling for Unsupervised Event-based Semantic Segmentation

arXiv:2403.16788v1 Announce Type: new
Abstract: 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 …

abstract arxiv cameras capability cs.cv data deal event hybrid image labeling labels lighting segmentation semantic speed type unsupervised