Quantifying and detecting collective motion by manifold learning

The analysis of collective motion has attracted many researchers in artificial intelligence. Though plenty of works have been done on this topic, the achieved performance is still unsatisfying due to the complex nature of collective motions. By investigating the similarity of individuals, this paper proposes a novel framework for both quantifying and detecting collective motions. Our main contributions are threefold: (1) the time-varying dynamics of individuals are deeply investigated to better characterize the individual motion; (2) a structure-based collectiveness measurement is designed to precisely quantify both individual-level and scene-level properties of collective motions; (3) a multi-stage clustering strategy is presented to discover a more comprehensive understanding of the crowd scenes, containing both local and global collective motions. Extensive experimental results on real world data sets show that our method is capable of handling crowd scenes with complicated structures and various dynamics, and demonstrate its superior performance against state-of-the-art competitors.