[00146733]Multi-complementary features adaptive fusion based on game theory for robust visual object tracking

To fully develop the complementary advantages of different visual features and improve the robustness of multi-feature fusion, we propose a robust correlation filter tracker with multi-complementary features adaptive fusion based on game theory. By combining the complementary features selected from handcrafted features and convolution features, our method constructs two robust combined features in the tracking framework of Discriminative Correlation Filter （DCF）. Besides, by using the idea of game theory, the two combined features are regarded as the two sides of the game, and the best balance is achieved through continuous gaming in the tracking process, to obtain a more robust fused feature. The experimental results obtained on the OTB2015 benchmark demonstrate that our tracker improves the robustness of object tracking in complex scenarios such as occlusion and deformation, and performs favorable performance against other outstanding methods。