RPT++: Customized Feature Representation for Siamese Visual Tracking
This addresses a specific bottleneck in visual tracking for computer vision applications, offering incremental improvements over existing methods.
The paper tackles feature misalignment between classification and regression tasks in visual tracking by proposing two customized feature extractors, polar pooling and extreme pooling, which when integrated into the RPT tracker achieve new state-of-the-art performances on benchmarks like OTB-100, VOT2018, VOT2019, GOT-10k, TrackingNet, and LaSOT.
While recent years have witnessed remarkable progress in the feature representation of visual tracking, the problem of feature misalignment between the classification and regression tasks is largely overlooked. The approaches of feature extraction make no difference for these two tasks in most of advanced trackers. We argue that the performance gain of visual tracking is limited since features extracted from the salient area provide more recognizable visual patterns for classification, while these around the boundaries contribute to accurately estimating the target state. We address this problem by proposing two customized feature extractors, named polar pooling and extreme pooling to capture task-specific visual patterns. Polar pooling plays the role of enriching information collected from the semantic keypoints for stronger classification, while extreme pooling facilitates explicit visual patterns of the object boundary for accurate target state estimation. We demonstrate the effectiveness of the task-specific feature representation by integrating it into the recent and advanced tracker RPT. Extensive experiments on several benchmarks show that our Customized Features based RPT (RPT++) achieves new state-of-the-art performances on OTB-100, VOT2018, VOT2019, GOT-10k, TrackingNet and LaSOT.