lenet_5¶

lucid.models.lenet_5(**kwargs) → LeNet¶

Overview¶

The lenet_5 function returns an instance of the LeNet-5 architecture, a classic convolutional neural network designed for handwritten digit recognition and other small-scale image classification tasks.

It is suitable for grayscale input images with a size of \(32 \times 32\), such as those in the original LeNet dataset.

Total Parameters: 61,706

Function Signature¶

@register_model
def lenet_5(**kwargs) -> LeNet

Returns¶

LeNet: An instance of the LeNet class configured for the LeNet-5 architecture.

Architecture¶

The LeNet-5 architecture consists of:

Feature Extractor: Two convolutional layers: - Conv1: 1 input channel, 6 output channels, \(5 \times 5\) kernel, stride 1. - Conv2: 6 input channels, 16 output channels, \(5 \times 5\) kernel, stride 1.

Two average pooling layers with a \(2 \times 2\) kernel and stride 2.
Classifier: Three fully connected layers: - FC1: Input features: \(16 \times 5 \times 5 = 400\), output features: 120. - FC2: Input features: 120, output features: 84. - FC3: Input features: 84, output features: 10 (number of classes)

Example Usage¶

Below is an example of using the lenet_5 function to create a LeNet-5 model:

import lucid.models as models

# Create the LeNet-5 model
model = models.lenet_5()

# Create a sample input tensor (e.g., 32x32 grayscale image)
input_tensor = lucid.Tensor([...])

# Perform a forward pass
output = model(input_tensor)
print(output)