The paper proposes a 3D multi-object tracking (MOT) solution that leverages 2D detections from monocular cameras to automatically initiate, terminate, and re-identify tracks, as well as handle occlusions. The approach is based on a Bayesian multi-object formulation that integrates track initiation/termination, re-identification, occlusion handling, and data association into a single Bayes filtering recursion. To address the numerical intractability of the exact filter due to the exponentially growing number of terms, the authors develop an efficient approximation that incorporates object features and kinematics into the measurement model, improving data association and reducing the number of terms. The proposed solution is evaluated on challenging datasets, demonstrating significant improvements and robustness compared to existing multi-view MOT solutions, especially when camera configurations change dynamically. The source code is publicly available at https://github.com/linh-gist/mv-gmb-ab.The paper proposes a 3D multi-object tracking (MOT) solution that leverages 2D detections from monocular cameras to automatically initiate, terminate, and re-identify tracks, as well as handle occlusions. The approach is based on a Bayesian multi-object formulation that integrates track initiation/termination, re-identification, occlusion handling, and data association into a single Bayes filtering recursion. To address the numerical intractability of the exact filter due to the exponentially growing number of terms, the authors develop an efficient approximation that incorporates object features and kinematics into the measurement model, improving data association and reducing the number of terms. The proposed solution is evaluated on challenging datasets, demonstrating significant improvements and robustness compared to existing multi-view MOT solutions, especially when camera configurations change dynamically. The source code is publicly available at https://github.com/linh-gist/mv-gmb-ab.