jac99/MinkLoc3Dv2
MinkLoc3Dv2: Improving Point Cloud Based Place Recognition with Ranking-based Loss and Large Batch Training
This project helps autonomous vehicles, robots, and mapping systems recognize where they are by matching current 3D point cloud scans to a database of known locations. It takes raw 3D point cloud data as input and outputs a highly discriminative descriptor that accurately identifies the location. Robotics engineers, autonomous driving researchers, and anyone developing spatial recognition systems would use this to improve their system's ability to localize itself in complex environments.
No commits in the last 6 months.
Use this if you need to build or enhance a system for robust 3D place recognition using point cloud data, especially in large-scale or challenging environments.
Not ideal if your application relies on 2D image data for localization, or if computational efficiency on highly resource-constrained edge devices is your absolute top priority without specialized hardware.
Stars
94
Forks
17
Language
Python
License
MIT
Category
Last pushed
Jan 31, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/jac99/MinkLoc3Dv2"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
patrick-llgc/Learning-Deep-Learning
Paper reading notes on Deep Learning and Machine Learning
magicleap/SuperGluePretrainedNetwork
SuperGlue: Learning Feature Matching with Graph Neural Networks (CVPR 2020, Oral)
eric-yyjau/pytorch-superpoint
Superpoint Implemented in PyTorch: https://arxiv.org/abs/1712.07629
lucasb-eyer/pydensecrf
Python wrapper to Philipp Krähenbühl's dense (fully connected) CRFs with gaussian edge potentials.
changhao-chen/deep-learning-localization-mapping
A collection of deep learning based localization models