Towards Segmenting Anything That Moves

Detecting and segmenting individual objects, regardless of their category, is crucial for many applications such as action detection or robotic interaction. While this problem has been well-studied under the classic formulation of spatio-temporal grouping, state-of-the-art approaches do not make use of learning-based methods. To bridge this gap, we propose a simple learning-based approach for spatio-temporal grouping. Our approach leverages motion cues Irom optical 'low as a bottom-up signal Ibr separating ohject.sfrom each other Motion cues are then combined with appearance cues that provide a generic objectness prior for capturing the fill extent of objects. We show that our approach outperforms all prior work on the benchmark FBMS dataset. One potential worn with learning-based methods is that they might overfit to the particular type of objects that they have been trained on. 70 address this concern, we propose Iwo new benchmarks /br generic, moving object detection, and show that our model matches top down methods on common categories, life significantly out-performing both lop-down and bottom-up methods on never-belbre-seen categories.