DAL -- A Deep Depth-aware Long-term Tracker
This work addresses the problem of efficient and accurate RGBD tracking for computer vision applications, representing an incremental improvement by integrating depth into existing methods.
The authors tackled the trade-off between accuracy and speed in RGBD tracking by proposing a deep depth-aware long-term tracker that reformulates deep discriminative correlation filters to embed depth information, achieving state-of-the-art performance on benchmarks like Princeton RGBD, STC, and CDTB with a speed of 20 fps.
The best RGBD trackers provide high accuracy but are slow to run. On the other hand, the best RGB trackers are fast but clearly inferior on the RGBD datasets. In this work, we propose a deep depth-aware long-term tracker that achieves state-of-the-art RGBD tracking performance and is fast to run. We reformulate deep discriminative correlation filter (DCF) to embed the depth information into deep features. Moreover, the same depth-aware correlation filter is used for target re-detection. Comprehensive evaluations show that the proposed tracker achieves state-of-the-art performance on the Princeton RGBD, STC, and the newly-released CDTB benchmarks and runs 20 fps.