VirtuaLearn3D++ framework’s documentation

Contents:

• Introduction
• Data mining
  • Geometric features miner
  • Geometric features miner ++
  • Height features miner
  • Height features miner ++
  • HSV from RGB miner
  • Smooth features miner
  • Smooth features miner ++
  • Recount miner
  • Recount miner ++
  • Naive change miner
  • Take closest miner
  • Decorators
    • Furthest point sampling decorator
    • Minimum distance decimator decorator
    • Simple smoother decorator
• Machine learning
  • Models
    • Random forest classifier
    • Random forest C++ classifier
  • Training strategies
    • Base training
    • Auto-validation
    • Stratified K-folding
  • Training data pipelines
    • Class-wise sampler
    • Synthetic minority oversampling technique (SMOTE)
  • Hyperparameter tuning
    • Grid search
    • Random search
  • Decorators
    • Furthest point sampling decorator
    • Minimum distance decimator decorator
  • Working example
    • Training pipeline
    • Predictive pipeline
• Deep learning
  • Models
    • PointNet-based point-wise classifier
    • Hierarchical autoencoder point-wise classifier
    • Sparse 3D convolutional point-wise classifier
    • Transformed Octo-Random Forest point-wise classifier
  • Receptive fields
    • Grid
    • Furthest point sampling
    • Furthest point sampling++
    • Hierarchical furthest point sampling
    • Hierarchical furthest point sampling++
    • Hierarchical sparse grid
  • Optimizers
  • Losses
  • Sequencers and data augmentation
    • Deep learning sequencer
    • Sparse sequencer
    • Offline sequencer
  • Training paradigms
    • Continual learning
    • Transfer learning
    • Freezing layers
    • TORF training tools
  • Working examples
    • PointNet-like model
    • PointNet++-based model
    • KPConv-based model
    • SFL-NET-like model
    • PointTransformer-based model
    • GroupedPointTransformer-based model
    • PointMLP-based model
    • KPConvX-based model
    • ContextNet-based model
• Transformers
  • Class transformers
    • Class reducer
    • Class setter
    • Distance reclassifier
    • Directional reclassifier
    • Taut string reclassifier
  • Feature transformers
    • Minmax normalizer
    • Standardizer
    • Variance selector
    • K-Best selector
    • Percentile selector
    • Explicit selector
    • PCA transformer
  • Point transformers
    • Point cloud sampler
    • Simple structure smoother++
• Imputers
  • Removal imputer
  • Univariate imputer
• Clustering
  • Clusterers
    • DBScan clusterer
    • Bivariate critical clusterer
    • Simple features clusterer
    • Simple curve extractor
  • Post-processing
    • ClusterEnveloper
    • ClusterMarker
    • ClusterSelector
  • Decorators
    • Furthest point sampling decorator
    • Minimum distance decimator decorator
• Writers
  • Writer
  • Model writer
  • Predictions writer
  • Classified point cloud writer
  • Predictive pipeline writer
• Evaluators
  • Classification evaluator
  • Advanced classification evaluator
  • Classification uncertainty evaluator
  • TORF RF vs NN evaluator
  • Regression evaluator
  • Simple curve evaluator
  • Deep learning model evaluator
  • Raster grid evaluator
  • Raster grid evaluator (PP)
    • Hillshade (Horn)
    • Local Relief Model (LRM)
    • Resampling filter
    • Visualization for Archaeological Topography (VAT)
    • Multi-Scale Relief Model (MSRM)
• Pipelines
  • Advanced input
    • Input point cloud concatenation
  • Sequential pipeline
    • Working example
  • Predictive pipeline
    • Working example
• Configuration
  • How configuration is loaded
  • io.yml – Input/Output
  • eval.yml – Evaluators
  • report.yml – Reports
  • mining.yml – Data Mining
  • model.yml – Models
    • All models
    • Machine learning
    • Deep learning
  • test.yml – Tests
  • logging.yml – Logging
  • Summary
• Examples
  • List of examples
    • Random forest for point-wise classification in the Hessigheim March2018 dataset
    • KPConv-like model for point-wise classification in the Hessigheim March2018 dataset
    • SFLNet-like model for point-wise aneurysm detection in medical 3D point clouds (IntrA dataset)
    • SFLNet-like model for point-wise segmentation of jaw teeth (Teeth3DS dataset)
    • Active learning with Random Forest for topo-bathymetric point clouds (Pielach 2024 dataset)
    • Deep learning with offline sequencer and data mining in the DALES dataset
    • Automatic extraction of change clusters

License:

• License
  • Documentation
  • Source code

src
npu

Indices and tables

• Index

• Module Index

• Search Page