🤖 AI Summary
This work addresses the challenge of unsupervised independent moving object (IMO) segmentation from event camera data—a problem where existing methods heavily rely on costly annotated datasets and struggle with arbitrary numbers of dynamic objects. The proposed method introduces, for the first time, a pseudo-label generation mechanism grounded in multi-view geometric constraints, integrated with spatiotemporal event stream modeling and motion-consistency regularization to enable end-to-end unsupervised training. Crucially, it requires no predefined object count or manual annotations, substantially reducing data dependency. Evaluated on the EVIMO benchmark, the approach achieves segmentation accuracy competitive with state-of-the-art supervised methods, while demonstrating superior robustness in high-speed and highly dynamic scenes. These results validate the feasibility and effectiveness of unsupervised learning for event-based visual understanding.
📝 Abstract
Event cameras are a novel type of biologically inspired vision sensor known for their high temporal resolution, high dynamic range, and low power consumption. Because of these properties, they are well-suited for processing fast motions that require rapid reactions. Although event cameras have recently shown competitive performance in unsupervised optical flow estimation, performance in detecting independently moving objects (IMOs) is lacking behind, although event-based methods would be suited for this task based on their low latency and HDR properties. Previous approaches to event-based IMO segmentation have been heavily dependent on labeled data. However, biological vision systems have developed the ability to avoid moving objects through daily tasks without being given explicit labels. In this work, we propose the first event framework that generates IMO pseudo-labels using geometric constraints. Due to its unsupervised nature, our method can handle an arbitrary number of not predetermined objects and is easily scalable to datasets where expensive IMO labels are not readily available. We evaluate our approach on the EVIMO dataset and show that it performs competitively with supervised methods, both quantitatively and qualitatively.
