What question did this study set out to answer?

This research aims to improve visual tracking through robust classification and accurate target state estimation guidelines.

April 3, 2020Open Access

SiamFC++: Towards Robust and Accurate Visual Tracking with Target Estimation Guidelines

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial demonstrates enhanced tracking performance in generic object tracking, indicating improved guidelines for target estimation.

This research aims to improve visual tracking through robust classification and accurate target state estimation guidelines.
Designed a Fully Convolutional Siamese tracker++ (SiamFC++) with multiple branches for classification and estimation.
Developed practical guidelines for target state estimation in tracker design.
Conducted extensive analyses and ablation studies to validate the effectiveness of the proposed guidelines.
Achieved a state-of-the-art AUC score of 75.4 on the TrackingNet dataset running over 90 FPS.
Demonstrated high performance across five benchmarks: OTB2015, VOT2018, LaSOT, GOT-10k, TrackingNet.
Proved the effectiveness of new guidelines in improving both tracking and generalization ability.

Xu et al. (2020) studied this question.

View Full Paper

Ask AI

Bookmark

Randomized trial demonstrates enhanced tracking performance in generic object tracking, indicating improved guidelines for target estimation.