point-cloud-classification

GPT4Vis: What Can GPT-4 Do for Zero-shot Visual Recognition?

Official Code for ICML 2021 paper "Revisiting Point Cloud Shape Classification with a Simple and Effective Baseline"

Codes for automatic point-cloud-to-BIM conversion

Linked Dynamic Graph CNN: Learning through Point Cloud by Linking Hierarchical Features

Code for "Rethinking the compositionality of point clouds through regularization in the hyperbolic space" (NeurIPS 2022)

Official code of paper "Efficient Joint Gradient Based Attack Against SOR Defense for 3D Point Cloud Classification" (MM'2020)

[MMSP 2023, PCS 2024]

PyTorch implementation of SimpleView from "Revisiting Point Cloud Classification with a Simple and Effective Baseline", Goyal et al. (2020)

Python SDK for running cloud-based 3D data workflows on the DataFre8 platform.

Directional PointNet: 3D Environmental Classification for Wearable Robots

This repository refers to my master's thesis in the Data Science at Sapienza University. The objective is to build a model in order to perform the task of 3D fragment matching. Given two 3D scans of fragments, the model must predict whether they are adjacent or not.

Code for MaxEnt: A maximum entropy-based optimal neighbor selection for multispectral airborne LiDAR point cloud classification

Code for MEHPool: A maximum entropy based homogeneous neighbor selection and hierarchical graph pooling module for point cloud classification.

Energy-based open-set active learning for 2D and 3D object classification.

LiDAR ground/non-ground classification using LAStools, Global Mapper, and a custom multi-scale Python ground filter.

Research on 3D point cloud classification with spatial hashing. Benchmarked on ModelNet40: 37.6× less memory than a 3D CNN baseline, at a cost of 10.4% accuracy

DeepEMO: A Multi-Indicator Convolutional Neural Network-based Evolutionary Multi-Objective Algorithm