Awesome GitHub Repositories3D Object Classification

PointNet este o arhitectură de deep learning concepută pentru a procesa și clasifica nori de puncte 3D brute direct, fără voxelizare. Oferă un sistem pentru clasificarea obiectelor 3D, framework-uri de segmentare semantică pentru partiționarea norilor în categorii și instrumente pentru vizualizarea formelor 3D. Proiectul utilizează o rețea de transformare pentru a alinia norii de puncte într-un spațiu de coordonate canonic și folosește agregarea bazată pe funcții simetrice pentru a condensa caracteristicile punctuale în vectori globali, indiferent de ordinea punctelor. De asemenea, dispune de o arhitectură de grupare multi-scară pentru a extrage caracteristici geometrice ierarhice la diferite scări spațiale. Sistemul include pipeline-uri de date pentru conversia datelor brute de parsare interioară 3D și a formelor în fișiere HDF5 pentru antrenare. Capabilitățile de evaluare acoperă metrici de acuratețe a segmentării și randarea norilor de puncte 3D clasificați greșit în imagini cu trei vederi pentru analiza erorilor.

PointNet++ is a deep learning framework designed for processing and classifying 3D point cloud data. It utilizes a hierarchical feature learning architecture to extract geometric patterns from sampled 3D point sets. The framework implements a variety of 3D analysis tools, including a point cloud classifier for categorizing objects based on spatial coordinates and surface normals, a semantic scene segmenter for labeling surfaces in large-scale environments, and a tool for 3D object part segmentation. The system covers a broad range of capabilities including geometric feature extraction, 3D da

MeshCNN is a general-purpose deep neural network for 3D triangular meshes, which can be used for tasks such as 3D shape classification or segmentation. This framework includes convolution, pooling and unpooling layers which are applied directly on the mesh edges.

3D Generative Adversarial Network

This repository contains the implementation of our papers related with O-CNN. The code is released under the MIT license.

OctNet uses efficient space partitioning structures (i.e. octrees) to reduce memory and compute requirements of 3D convolutional neural networks, thereby enabling deep learning at high resolutions.

PointConv: Deep Convolutional Networks on 3D Point Clouds. CVPR 2019 Wenxuan Wu, Zhongang Qi, Li Fuxin.

Created by Itai Lang, Asaf Manor, and Shai Avidan from Tel Aviv University.

3D/Volumetric Convolutional Neural Networks with Theano+Lasagne.

Created by Yutong Feng, Yifan Feng, Haoxuan You, Xibin Zhao, Yue Gao from Tsinghua University.

Volumetric CNN (Convolutional Neural Networks) for Object Classification on 3D Data, with Torch implementation.

Voxel-Based Variational Autoencoders, VAE GUI, and Convnets for Classification

Here you can find the code for the BMVC 2017 version of "Orientation-boosted Voxel Nets for 3D Object Recognition", a.k.a ORION.

Paper: Feedback Networks, CVPR 2017.

SpiderCNN: Deep Learning on Point Sets with Parameterized Convolutional Filters. ECCV 2018 Yifan Xu, Tianqi Fan, Mingye Xu, Long Zeng, Yu Qiao.

1. Python (with necessary common libraries such as numpy, scipy, etc.) 2. TensorFlow 3. You need to prepare your data in *.mat file with the following format: - 'points': N x 3 array (x, y, z coordinates of the point cloud) - 'labels': N x 1 array (1-based integer per-point labels) - 'category':…

Field Probing Neural Networks for 3D Data