GAKP: GRU Association and Kalman Prediction for Multiple Object Tracking
This work provides a more efficient and robust online MOT algorithm for applications like video surveillance and smart cities, particularly beneficial when large amounts of training data are unavailable.
This paper addresses the challenge of modeling long-term temporal dependencies in Multiple Object Tracking (MOT) with limited training data. The proposed GAKP method integrates auto-tuning Kalman prediction with a Gated Recurrent Unit (GRU), achieving competitive performance on challenging MOT benchmarks while being faster and more robust than state-of-the-art RNN-based online MOT algorithms.
Multiple Object Tracking (MOT) has been a useful yet challenging task in many real-world applications such as video surveillance, intelligent retail, and smart city. The challenge is how to model long-term temporal dependencies in an efficient manner. Some recent works employ Recurrent Neural Networks (RNN) to obtain good performance, which, however, requires a large amount of training data. In this paper, we proposed a novel tracking method that integrates the auto-tuning Kalman method for prediction and the Gated Recurrent Unit (GRU) and achieves a near-optimum with a small amount of training data. Experimental results show that our new algorithm can achieve competitive performance on the challenging MOT benchmark, and faster and more robust than the state-of-the-art RNN-based online MOT algorithms.