If you have a pretrained model in mxnet *.json format.

Using the Fine-tuning strategy to transfer learned recognition capabilities from general domains to the specific challenge.

You can use the example code below directly in the R console.

For example for the DesNet.

  Dense_model = mx.model.load('model/densenet-imagenet-169-0', 125)

  all_layers = Dense_model$symbol$get.internals()
  relu1_output = which(all_layers$outputs == 'relu1_output') %>% all_layers$get.output()
  softmax_output = which(all_layers$outputs == 'softmax_output') %>% all_layers$get.output()
  
  out = mx.symbol.Group(c(relu1_output, softmax_output))
  executor = mx.simple.bind(symbol = out, data = c(224, 224, 3, 1), ctx = mx.cpu())
  
  mx.exec.update.arg.arrays(executor, Dense_model$arg.params, match.name = TRUE)
  mx.exec.update.aux.arrays(executor, Dense_model$aux.params, match.name = TRUE)

And run the executor or train by other mxnet function.

Using symbol.get_internals to get the internal parts, can only get symbol from start.

You may rewrite the whole Net.

Or you may use function in '1. LeNet/3. convert json to symbol.R', converting json files to R code for the example of LeNet.

Also, there is an example of DesNet.

The example code in here'1. LeNet/3. convert json to symbol.R'

Pre-Train densenet model is downloading from

• DenseNet