📌 Overview
This project implements and evaluates a deep learning–based 3D point cloud classification framework using the ModelNet++ architecture. The work focuses on learning discriminative features directly from raw point cloud data and performing object classification on the ModelNet dataset. The architecture follows the hierarchical feature learning strategy inspired by PointNet++. This work is perfomed during the time interval of my intership Ank Computing.
- Implement ModelNet++ architecture for 3D object classification
- Train the network on ModelNet dataset
- Analyze training and validation performance
- Evaluate classification accuracy and loss behavior
- The dataset used ModelNet10
https://github.com/yanx27/Pointnet_Pointnet2_pytorch/tree/master?tab=readme-ov-file#pytorch-implementation-of-pointnet-and-pointnet
PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space Charles R. Qi Li Yi HaoSu LeonidasJ.Guibas Stanford University