-
Notifications
You must be signed in to change notification settings - Fork 0
Home
Connor Novak edited this page Dec 13, 2019
·
7 revisions
In this project, we attempted to use machine learning to detect and classify different objects in a 3D pointcloud using the labeled datasets from Waymo. Our process and observations are outlined in this wiki, and we recommend navigating this wiki by going through the pages in this order:
- Home
- Data Sources
- PCL Segmentation
- PCL Feature Representation
- Model Development
On each page, the next page should be linked at the bottom.
- Medium: "3D Object Detection from Lidar Data with Machine Learning”
- Stackoverflow: "Clustering with an unknown number of clusters"
- Human-centric Computing and Information Sciences: “Classifying 3D objects in LiDAR point clouds with a back-propagation neural network”
- ResearchGate: "A Comparative Study of Machine Learning Regression Models on Lidar Data"
- "A comparison of various machine learning techniques for Aerial LiDAR Data Classification using Support Vector Machines (SVM)"
- Velodyne: "LIDAR-based 3D Object Perception"
- Human-centric Computing and Information Sciences: "A 3D localisation method in indoor environments for virtual reality applications"
- Scikitlearn: “Choosing the Right Estimator” Map
- Waymo: Open Dataset
- Scikitlearn: Different clustering algorithms https://scikit-learn.org/stable/modules/clustering.html