jomariya23156 / superglue-for-visual-place-recognition

Using SuperGlue (from Magic Leap team) in Visual Place Recognition tasks. Providing full workflow from videos/images to end-to-end API and step-by-step how to use all codes.

It's me, JOM. 👋

📚 A passionate well-rounded ML Engineer focusing on MLOps and ML Solution Architecture 📚

• 🚀 Open to work: Seeking International Opportunities. Open to Relocation Globally. :)

• 🌱 Expertise: Computer Vision, Medical Imaging, ML System Design, ML Solution Architecture

• ⚡ Check this out! jomariya23156.github.io

• 📫 Reach me [email protected]

In case you had a crush on me:

Beat My Tech Stack:

         

 

superglue-for-visual-place-recognition's Issues

Does this repo apply to the pitts dataset

Hi,
Does this repo apply to the pittsburgh dataset, and how to use your code with the pittsburgh dataset
thank you

ranking score

Hi, thanks for your work.

I want to ask why the ranking score uses np.multiply(matches,match_confidence)? Here matches should be the indexes of the matches, is it meaningful to multiply it with the matching confidence?