GESS: Multi-cue Guided Local Feature Learning via Geometric and Semantic Synergy

Yang Yi, Xieyuanli Chen, Jinpu Zhang^*, Hui Shen^*, Dewen Hu

Project Website (Coming Soon)

Benchmarking Results

1. HPatches Image Matching Benchmark

2. Visual Localization Benchmark

aachen-v1.0

aachen-v1.1

3. Relative pose estimation

For detailed experimental results, please refer to:

• 📊 Detailed Results

4. 3D Reconstruction

For detailed experimental results, please refer to:

• 📊 Detailed Results