Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Object recognition: apply *h5 cnn model to find an specific handwritten digit #17

Open
Leprechault opened this issue Jan 7, 2020 · 1 comment

Comments

@Leprechault
Copy link

I don't find any material about object detection using *h5 cnn adjusted model.
I try to create a simple example because I don't find in any book/web, I've like to apply any object detection technique to find handwritten digits in an image with several digits (wh_img). First, a create a CNN using the classic MNIST example:

Preparing the Data - using MNIST dataset is included with Kera

library(keras)
mnist <- dataset_mnist()
x_train <- mnist$train$x
y_train <- mnist$train$y
x_test <- mnist$test$x
y_test <- mnist$test$y

#reshape
x_train <- array_reshape(x_train, c(nrow(x_train), 784))
x_test <- array_reshape(x_test, c(nrow(x_test), 784))
#rescale
x_train <- x_train / 255
x_test <- x_test / 255

y_train <- to_categorical(y_train, 10)
y_test <- to_categorical(y_test, 10)

CNN model

model <- keras_model_sequential()
model %>%
layer_dense(units = 256, activation = 'relu', input_shape = c(784)) %>%
layer_dropout(rate = 0.4) %>%
layer_dense(units = 128, activation = 'relu') %>%
layer_dropout(rate = 0.3) %>%
layer_dense(units = 10, activation = 'softmax')

model %>% compile(
loss = 'categorical_crossentropy',
optimizer = optimizer_rmsprop(),
metrics = c('accuracy')
)

#Training and Evaluation
history <- model %>% fit(
x_train, y_train,
epochs = 30, batch_size = 128,
validation_split = 0.2
)
plot(history)

Save the model

model %>% save_model_hdf5("cnn_digits.h5")

Now, I've to apply any object detection technique in my y.png image with several handwritten digits to find only 4 numerals using my trained model :

Open an image with handwrite numbers collections

y = "http://mariakravtsova.us/img/numbers.png"
download.file(y,'y.png', mode = 'wb')
library("png")
wh_img <- readPNG("y.png")
plot.new()
rasterImage(wh_img,0,0,1,1)

library(image.darknet)

Try to use *h5 model to object detection only 4 handwritten digits

yolo_digits <- image_darknet_model(type = ‘detect’, labels ="4", model = "cnn_digits.h5")

x <- image_darknet_detect(file = "/Documents/numbers.png",
object = yolo_digits,
threshold = 0.4)

This is possible in R? Because in Python there are a lot of functions, choose the window size, etc. but in R not :(

@jwijffels
Copy link
Contributor

jwijffels commented Jan 8, 2020

You have build a model with keras and saved it as cnn_digits.h5. You can not pass such a model to work with darknet as that requires a .cfg file. Keras and darknet are 2 different frameworks.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants