A0308
Title: Variable target scalable particle filter
Authors: Ning Ning - Texas A&M University (United States) [presenting]
Abstract: The challenge of tracking a variable number of interacting targets is addressed. The primary goal is to accurately detect targets entering and leaving the scene while maintaining a precise trajectory record for each target throughout their presence. Developing methods that effectively handle this complexity is vital for scenarios involving the continuous tracking of numerous interacting targets. To tackle this problem, the variable target scalable particle filter (VTSPF) is introduced within an online learning framework. VTSPF efficiently tracks multiple moving targets exhibiting complex interactions. Importantly, it demonstrates scalability in both spatial and temporal dimensions.
