DFAT

The official implementation of the Information Fusion paper "Exploring Fusion Strategies for Accurate RGBT Visual Object Tracking".

It is supported by the PRCI-Lab (https://github.com/PRCI-Lab)

🫵Find our survey work at another repo

Introduction

Based on the pysot, DFAT mainly focus on the bias caused by the different imaging mechanism.

Illustration of databias

Download models

The used model can be downloaded from baidu_disk. CODE:nfbe

Download testing datasets

Testing dataset like GTOT, RGBT210, RGBT234 can be downloaded from Chenglong Li's website

VOT-RGBT2019 and VOT-RGBT2020 datasets can be downloaded from baidu_disk. CODE:bcyo

Eval tracker

For VOT toolkit: python_siam_base.py

For testing: python test_gtot.py/test_vot.py

RFN

The RFN block is introduced from RFN-Nest

Contributors

• Zhangyong Tang