DEFT: Detection Embeddings for Tracking

3 Feb 2021  ·  Mohamed Chaabane, Peter Zhang, J. Ross Beveridge, Stephen O'Hara ·

Most modern multiple object tracking (MOT) systems follow the tracking-by-detection paradigm, consisting of a detector followed by a method for associating detections into tracks. There is a long history in tracking of combining motion and appearance features to provide robustness to occlusions and other challenges, but typically this comes with the trade-off of a more complex and slower implementation. Recent successes on popular 2D tracking benchmarks indicate that top-scores can be achieved using a state-of-the-art detector and relatively simple associations relying on single-frame spatial offsets -- notably outperforming contemporary methods that leverage learned appearance features to help re-identify lost tracks. In this paper, we propose an efficient joint detection and tracking model named DEFT, or "Detection Embeddings for Tracking." Our approach relies on an appearance-based object matching network jointly-learned with an underlying object detection network. An LSTM is also added to capture motion constraints. DEFT has comparable accuracy and speed to the top methods on 2D online tracking leaderboards while having significant advantages in robustness when applied to more challenging tracking data. DEFT raises the bar on the nuScenes monocular 3D tracking challenge, more than doubling the performance of the previous top method. Code is publicly available.

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Multi-Object Tracking KITTI Tracking test DEFT HOTA 74.23 # 3
Multiple Object Tracking KITTI Tracking test DEFT MOTA 88.95 # 5
Multi-Object Tracking MOT16 DEFT MOTA 68.03 # 11
Multi-Object Tracking MOT17 DEFT MOTA 66.6 # 25
3D Multi-Object Tracking nuScenes DEFT AMOTA 0.18 # 102

Methods