Conv2DTranspose is a Keras layer that learns to upsample feature maps, restoring spatial resolution after convolutional layers compress an image. section 11.3.2
Expanding a segmentation feature map
A feature map of shape (25, 25, 256) can pass through corresponding Conv2DTranspose layers before a final Conv2D layer produces a (200, 200, 3) pixel-class output. With strides = 2 and padding = "same", a (50, 50, 128) map can return to (100, 100, 64). section 11.3.2
A segmentation model typically uses Conv2DTranspose layers to:
- Expand a compact feature map back toward the input image dimensions. section 11.3.2
- Follow a stack of strided Conv2D layers with corresponding upsampling layers. section 11.5
- Support per-pixel predictions by feeding the expanded representation to a final Conv2D layer. section 11.3.2
get_model <- function(img_size, num_classes) {
conv <- function(..., padding = "same", activation = "relu") #1
layer_conv_2d(..., padding = padding, activation = activation)
conv_transpose <- function(..., padding = "same", activation = "relu")
layer_conv_2d_transpose(..., padding = padding, activation = activation)
input <- keras_input(shape = c(img_size, 3))
output <- input |>
layer_rescaling(scale = 1/255) |> #2
conv(64, 3, strides = 2) |>
conv(64, 3) |>
conv(128, 3, strides = 2) |>
conv(128, 3) |>
conv(256, 3, strides = 2) |>
conv(256, 3) |>
conv_transpose(256, 3) |>
conv_transpose(256, 3, strides = 2) |>
conv_transpose(128, 3) |>
conv_transpose(128, 3, strides = 2) |>
conv_transpose(64, 3) |>
conv_transpose(64, 3, strides = 2) |>
conv(num_classes, 3, activation = "softmax") #3
keras_model(input, output)
}
model <- get_model(img_size = img_size, num_classes = 3)
model
Model: "functional"
┌─────────────────────────────────┬────────────────────────┬───────────────┐
│ Layer (type) │ Output Shape │ Param # │
├─────────────────────────────────┼────────────────────────┼───────────────┤
│ input_layer (InputLayer) │ (None, 200, 200, 3) │ 0 │
├─────────────────────────────────┼────────────────────────┼───────────────┤
│ rescaling (Rescaling) │ (None, 200, 200, 3) │ 0 │
├─────────────────────────────────┼────────────────────────┼───────────────┤
│ conv2d (Conv2D) │ (None, 100, 100, 64) │ 1,792 │
├─────────────────────────────────┼────────────────────────┼───────────────┤
│ conv2d_1 (Conv2D) │ (None, 100, 100, 64) │ 36,928 │
├─────────────────────────────────┼────────────────────────┼───────────────┤
│ conv2d_2 (Conv2D) │ (None, 50, 50, 128) │ 73,856 │
├─────────────────────────────────┼────────────────────────┼───────────────┤
│ conv2d_3 (Conv2D) │ (None, 50, 50, 128) │ 147,584 │
├─────────────────────────────────┼────────────────────────┼───────────────┤
│ conv2d_4 (Conv2D) │ (None, 25, 25, 256) │ 295,168 │
├─────────────────────────────────┼────────────────────────┼───────────────┤
│ conv2d_5 (Conv2D) │ (None, 25, 25, 256) │ 590,080 │
└─────────────────────────────────┴────────────────────────┴───────────────┘
┌─────────────────────────────────┬────────────────────────┬───────────────┐
│ conv2d_transpose │ (None, 25, 25, 256) │ 590,080 │
│ (Conv2DTranspose) │ │ │
├─────────────────────────────────┼────────────────────────┼───────────────┤
│ conv2d_transpose_1 │ (None, 50, 50, 256) │ 590,080 │
│ (Conv2DTranspose) │ │ │
├─────────────────────────────────┼────────────────────────┼───────────────┤
│ conv2d_transpose_2 │ (None, 50, 50, 128) │ 295,040 │
│ (Conv2DTranspose) │ │ │
├─────────────────────────────────┼────────────────────────┼───────────────┤
│ conv2d_transpose_3 │ (None, 100, 100, 128) │ 147,584 │
│ (Conv2DTranspose) │ │ │
├─────────────────────────────────┼────────────────────────┼───────────────┤
│ conv2d_transpose_4 │ (None, 100, 100, 64) │ 73,792 │
│ (Conv2DTranspose) │ │ │
├─────────────────────────────────┼────────────────────────┼───────────────┤
│ conv2d_transpose_5 │ (None, 200, 200, 64) │ 36,928 │
│ (Conv2DTranspose) │ │ │
├─────────────────────────────────┼────────────────────────┼───────────────┤
│ conv2d_6 (Conv2D) │ (None, 200, 200, 3) │ 1,731 │
└─────────────────────────────────┴────────────────────────┴───────────────┘
Total params: 2,880,643 (10.99 MB)
Trainable params: 2,880,643 (10.99 MB)
Non-trainable params: 0 (0.00 B)
Runnable R/Keras model code shows Conv2DTranspose layers alternating ordinary and stride-2 upsampling stages before a softmax segmentation output. section 11.3.2
