EnsembleMOT: A Step towards Ensemble Learning of Multiple Object Tracking
This work addresses the complexity of multiple object tracking for computer vision applications, but it is incremental as it applies ensemble learning to an existing task.
The paper tackles the problem of multiple object tracking by proposing EnsembleMOT, an ensemble method that merges results from various trackers with spatio-temporal constraints, achieving improved performance on the MOT17 dataset.
Multiple Object Tracking (MOT) has rapidly progressed in recent years. Existing works tend to design a single tracking algorithm to perform both detection and association. Though ensemble learning has been exploited in many tasks, i.e, classification and object detection, it hasn't been studied in the MOT task, which is mainly caused by its complexity and evaluation metrics. In this paper, we propose a simple but effective ensemble method for MOT, called EnsembleMOT, which merges multiple tracking results from various trackers with spatio-temporal constraints. Meanwhile, several post-processing procedures are applied to filter out abnormal results. Our method is model-independent and doesn't need the learning procedure. What's more, it can easily work in conjunction with other algorithms, e.g., tracklets interpolation. Experiments on the MOT17 dataset demonstrate the effectiveness of the proposed method. Codes are available at https://github.com/dyhBUPT/EnsembleMOT.