Convolutional Recurrent Predictor: Implicit Representation for Multi-target Filtering and Tracking.

Defining a multi-target motion model, an important step of tracking algorithms, is a challenging task due to various factors, from its theoretical formulation to its computational complexity. Using fixed models (as in several generative Bayesian algorithms, such as Kalman filters) can fail to accurately predict sophisticated target motions. On the other hand, sequential learning of the motion mode...