SiamFDA: feature dynamic activation siamese network for visual tracking

Abstract In this paper, we present a novel anchor-free visual tracking framework, referred to as feature dynamic activation siamese network (SiamFDA), which addresses the issue of ignoring global spatial information in current Siamese network-based tracking algorithms. Our approach captures long-ran...

Ausführliche Beschreibung

Bibliographische Detailangaben
Hauptverfasser: Jialiang Gu, Ying She, Yi Yang
Format: Artikel
Sprache:English
Veröffentlicht: Nature Portfolio 2024-02-01
Schriftenreihe:Scientific Reports
Online Zugang:https://doi.org/10.1038/s41598-024-55487-7