Depth Masked Discriminative Correlation Filter
This improves object tracking for robotics and AR/VR applications by handling occlusions more effectively, though it is an incremental enhancement of existing DCF methods.
The paper tackles object tracking in RGB-D videos by proposing DM-DCF, which uses depth information for occlusion detection and masking. The method achieves state-of-the-art results on the Princeton RGBD Tracking Benchmark and runs significantly faster than competitors.
Depth information provides a strong cue for occlusion detection and handling, but has been largely omitted in generic object tracking until recently due to lack of suitable benchmark datasets and applications. In this work, we propose a Depth Masked Discriminative Correlation Filter (DM-DCF) which adopts novel depth segmentation based occlusion detection that stops correlation filter updating and depth masking which adaptively adjusts the spatial support for correlation filter. In Princeton RGBD Tracking Benchmark, our DM-DCF is among the state-of-the-art in overall ranking and the winner on multiple categories. Moreover, since it is based on DCF, ``DM-DCF`` runs an order of magnitude faster than its competitors making it suitable for time constrained applications.