Robust Visual Localization via Semantic-Guided Multi-Scale Transformer

Zhongtao Tian, Wenhao Huang, Zhidong Chen +1

Visual localization remains challenging in dynamic environments where fluctuating lighting, adverse weather, and moving objects disrupt appearance cues. Despite advances in feature…