GLAD: Generative Language-Assisted Visual Tracking for Low-Semantic Templates

Open paper

Benchmark	Model	Rank	Results
visual-tracking-on-tnl2k	GLAD-B256	–	AUC: 59.7precision: 62.2
visual-tracking-on-tnl2k	GLAD-L384	–	AUC: 62.1precision: 66.2