Hacker News new | past | comments | ask | show | jobs | submit login

In my practical experience, the majority computer graphics and vision frameworks (e.g. point cloud processing in Point Cloud Library, feature matching in colmap) all use FLANN (https://www.cs.ubc.ca/research/flann/) which employs a variety of NN algorithms that specialize in high dimensional spaces (e.g. 128 dimensional SIFT descriptors).



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: