DINO-Mix: Enhancing Visual Place Recognition with Foundational Vision Model and Feature Mixing

1 code implementation • 1 Nov 2023 • Gaoshuang Huang, Yang Zhou, Xiaofei Hu, Chenglong Zhang, Luying Zhao, Wenjian Gan, Mingbo Hou

In this study, we utilize the DINOv2 model as the backbone network for trimming and fine-tuning to extract robust image features.

Visual Place Recognition