Table of Contents

Class MeshCNN<T>

Namespace
AiDotNet.NeuralNetworks
Assembly
AiDotNet.dll

Implements the MeshCNN architecture for processing 3D triangle meshes.

public class MeshCNN<T> : NeuralNetworkBase<T>, INeuralNetworkModel<T>, INeuralNetwork<T>, IFullModel<T, Tensor<T>, Tensor<T>>, IModel<Tensor<T>, Tensor<T>, ModelMetadata<T>>, IModelSerializer, ICheckpointableModel, IParameterizable<T, Tensor<T>, Tensor<T>>, IFeatureAware, IFeatureImportance<T>, ICloneable<IFullModel<T, Tensor<T>, Tensor<T>>>, IGradientComputable<T, Tensor<T>, Tensor<T>>, IJitCompilable<T>, IInterpretableModel<T>, IInputGradientComputable<T>, IDisposable

Type Parameters

T

The numeric type used for calculations (typically float or double).

Inheritance
MeshCNN<T>
Implements
IFullModel<T, Tensor<T>, Tensor<T>>
IModel<Tensor<T>, Tensor<T>, ModelMetadata<T>>
IParameterizable<T, Tensor<T>, Tensor<T>>
ICloneable<IFullModel<T, Tensor<T>, Tensor<T>>>
IGradientComputable<T, Tensor<T>, Tensor<T>>
Inherited Members
Extension Methods

Remarks

MeshCNN is a deep learning architecture that operates directly on 3D mesh data by treating edges as the fundamental unit of computation. This enables learning from the mesh structure itself rather than converting to voxels or point clouds.

For Beginners: MeshCNN processes 3D shapes represented as triangle meshes.

Key concepts:

  • Mesh: A 3D surface made of connected triangles (vertices + faces)
  • Edge: A line segment connecting two vertices, shared by up to 2 faces
  • Edge features: Properties like dihedral angle, edge ratios, face angles

How it works:

  1. Extract edge features from the mesh (5 features per edge by default)
  2. Apply edge convolutions to learn patterns in edge neighborhoods
  3. Pool edges by removing less important ones (simplifies the mesh)
  4. Repeat conv + pool to build hierarchical features
  5. Aggregate edge features for classification/segmentation

Applications:

  • 3D shape classification (e.g., recognize chair vs table)
  • Mesh segmentation (label each part of a 3D model)
  • Shape retrieval (find similar 3D models)

Reference: "MeshCNN: A Network with an Edge" by Hanocka et al., SIGGRAPH 2019

Constructors

MeshCNN()

Initializes a new instance of the MeshCNN<T> class with default options.

public MeshCNN()

Remarks

Creates a MeshCNN with default configuration suitable for ModelNet40 classification.

MeshCNN(MeshCNNOptions, IGradientBasedOptimizer<T, Tensor<T>, Tensor<T>>?, ILossFunction<T>?)

Initializes a new instance of the MeshCNN<T> class with specified options.

public MeshCNN(MeshCNNOptions options, IGradientBasedOptimizer<T, Tensor<T>, Tensor<T>>? optimizer = null, ILossFunction<T>? lossFunction = null)

Parameters

options MeshCNNOptions

Configuration options for the MeshCNN.

optimizer IGradientBasedOptimizer<T, Tensor<T>, Tensor<T>>

The optimizer for training. Defaults to Adam if null.

lossFunction ILossFunction<T>

The loss function. Defaults based on task type if null.

Remarks

For Beginners: Creates a MeshCNN with custom configuration.

Exceptions

ArgumentNullException

Thrown when options is null.

MeshCNN(int, int, ILossFunction<T>?)

Initializes a new instance of the MeshCNN<T> class with simple parameters.

public MeshCNN(int numClasses, int inputFeatures = 5, ILossFunction<T>? lossFunction = null)

Parameters

numClasses int

Number of output classes for classification.

inputFeatures int

Number of input features per edge. Default is 5.

lossFunction ILossFunction<T>

The loss function. Defaults based on task type if null.

Remarks

For Beginners: Creates a MeshCNN with default architecture settings.

Properties

ConvChannels

Gets the channel configuration for edge convolution layers.

public int[] ConvChannels { get; }

Property Value

int[]

InputFeatures

Gets the number of input features per edge.

public int InputFeatures { get; }

Property Value

int

NumClasses

Gets the number of output classes for classification.

public int NumClasses { get; }

Property Value

int

PoolTargets

Gets the pooling targets for mesh simplification.

public int[] PoolTargets { get; }

Property Value

int[]

Methods

Backward(Tensor<T>)

Performs backward pass through the network.

public Tensor<T> Backward(Tensor<T> outputGradient)

Parameters

outputGradient Tensor<T>

Gradient of loss with respect to output.

Returns

Tensor<T>

Gradient with respect to input.

CreateNewInstance()

Creates a new instance for cloning.

protected override IFullModel<T, Tensor<T>, Tensor<T>> CreateNewInstance()

Returns

IFullModel<T, Tensor<T>, Tensor<T>>

New MeshCNN instance.

Remarks

For Beginners: This creates a blank version of the same type of neural network.

It's used internally by methods like DeepCopy and Clone to create the right type of network before copying the data into it.

DeserializeNetworkSpecificData(BinaryReader)

Deserializes network-specific data.

protected override void DeserializeNetworkSpecificData(BinaryReader reader)

Parameters

reader BinaryReader

Binary reader.

Remarks

This method is called at the end of the general deserialization process to allow derived classes to read any additional data specific to their implementation.

For Beginners: Continuing the suitcase analogy, this is like unpacking that special compartment. After the main deserialization method has unpacked the common items (layers, parameters), this method allows each specific type of neural network to unpack its own unique items that were stored during serialization.

Forward(Tensor<T>)

Performs a forward pass through the network.

public Tensor<T> Forward(Tensor<T> input)

Parameters

input Tensor<T>

Edge features tensor with shape [numEdges, InputFeatures].

Returns

Tensor<T>

Classification logits with shape [NumClasses].

Remarks

Call SetEdgeAdjacency(int[,]) before calling this method.

Exceptions

InvalidOperationException

Thrown when edge adjacency is not set.

GetModelMetadata()

Gets metadata about this model.

public override ModelMetadata<T> GetModelMetadata()

Returns

ModelMetadata<T>

Model metadata.

InitializeLayers()

Initializes the layers of the MeshCNN network.

protected override void InitializeLayers()

Remarks

If the architecture provides custom layers, those are used. Otherwise, default layers are created using CreateDefaultMeshCNNLayers(NeuralNetworkArchitecture<T>, int, int[]?, int[]?, int[]?, int, bool, double, bool).

Predict(Tensor<T>)

Generates predictions for the given input.

public override Tensor<T> Predict(Tensor<T> input)

Parameters

input Tensor<T>

Edge features tensor.

Returns

Tensor<T>

Classification logits.

Remarks

For Beginners: This is the main method you'll use to get results from your trained neural network. You provide some input data (like an image or text), and the network processes it through all its layers to produce an output (like a classification or prediction).

SerializeNetworkSpecificData(BinaryWriter)

Serializes network-specific data.

protected override void SerializeNetworkSpecificData(BinaryWriter writer)

Parameters

writer BinaryWriter

Binary writer.

Remarks

This method is called at the end of the general serialization process to allow derived classes to write any additional data specific to their implementation.

For Beginners: Think of this as packing a special compartment in your suitcase. While the main serialization method packs the common items (layers, parameters), this method allows each specific type of neural network to pack its own unique items that other networks might not have.

SetEdgeAdjacency(int[,])

Sets the edge adjacency for the current mesh being processed.

public void SetEdgeAdjacency(int[,] edgeAdjacency)

Parameters

edgeAdjacency int[,]

A 2D array of shape [numEdges, NumNeighbors] containing neighbor edge indices.

Remarks

For Beginners: Before processing a mesh, you must tell the network how edges are connected. This method sets that connectivity information.

Call this method before each Forward pass with a new mesh.

Exceptions

ArgumentNullException

Thrown when edgeAdjacency is null.

Train(Tensor<T>, Tensor<T>)

Trains the network on a single batch.

public override void Train(Tensor<T> input, Tensor<T> expectedOutput)

Parameters

input Tensor<T>

Edge features tensor.

expectedOutput Tensor<T>

Ground truth labels.

Remarks

This method performs one training step on the neural network using the provided input and expected output. It updates the network's parameters to reduce the error between the network's prediction and the expected output.

For Beginners: This is how your neural network learns. You provide: - An input (what the network should process) - The expected output (what the correct answer should be)

The network then:

  1. Makes a prediction based on the input
  2. Compares its prediction to the expected output
  3. Calculates how wrong it was (the loss)
  4. Adjusts its internal values to do better next time

After training, you can get the loss value using the GetLastLoss() method to see how well the network is learning.

UpdateParameters(Vector<T>)

Updates network parameters using a flat parameter vector.

public override void UpdateParameters(Vector<T> parameters)

Parameters

parameters Vector<T>

Vector containing all parameters.

Remarks

For Beginners: During training, a neural network's internal values (parameters) get adjusted to improve its performance. This method allows you to update all those values at once by providing a complete set of new parameters.

This is typically used by optimization algorithms that calculate better parameter values based on training data.

Edit this page