The AzureVision package is an R frontend to Azure Computer Vision and Azure Custom Vision. These services let you leverage Microsoft’s Azure cloud to carry out visual recognition tasks using advanced image processing models, with minimal machine learning expertise.
The basic idea behind Custom Vision is to take a pre-built image recognition model supplied by Azure, and customise it for your needs by supplying a set of images with which to update it. All model training and prediction is done in the cloud, so you don’t need a powerful machine of your own. Similarly, since you are starting with a model that has already been trained, you don’t need a very large dataset or long training times to obtain good predictions (ideally).
This article walks you through how to create, train and deploy a Custom Vision model in R, using AzureVision.
Creating the resources
You can create the Custom Vision resources via the Azure portal, or in R using the facilities provided by AzureVision. Note that Custom Vision requires at least two resources to be created: one for training, and one for prediction. The available service tiers for Custom Vision are
F0 (free, limited to 2 projects for training and 10k transactions/month for prediction) and
S0. Here is the R code for creating the resources:
library(AzureVision) # insert your tenant, subscription, resgroup name and location here rg <- AzureRMR::get_azure_login(tenant)$ get_subscription(sub_id)$ create_resource_group(rg_name, location=rg_location) # insert your desired Custom Vision resource names here res <- rg$create_cognitive_service(custvis_resname, service_type="CustomVision.Training", service_tier="S0") pred_res <- rg$create_cognitive_service(custvis_predresname, service_type="CustomVision.Prediction", service_tier="S0")
Custom Vision defines two different types of endpoint: a training endpoint, and a prediction endpoint. Somewhat confusingly, they can both use the same hostname, but with different URL paths and authentication keys. To start, call the
customvision_training_endpoint function with the service URL and key.
Custom Vision is organised hierarchically. At the top level, we have a project, which represents the data and model for a specific task. Within a project, we have one or more iterations of the model, built on different sets of training images. Each iteration in a project is independent: you can create (train) an iteration, deploy it, and delete it without affecting other iterations.
In turn, there are three different types of projects:
- A multiclass classification project is for classifying images into a set of tags, or target labels. An image can be assigned to one tag only.
- A multilabel classification project is similar, but each image can have multiple tags assigned to it.
- An object detection project is for detecting which objects, if any, from a set of candidates are present in an image.
Let’s create a classification project:
Here, we specify the export target to be
standard to support exporting the final model to one of various standalone formats, eg TensorFlow, CoreML or ONNX. The default is
none, in which case the model stays on the Custom Vision server. The advantage of
none is that the model can be more complex, resulting in potentially better accuracy.
Adding and tagging images
Since a Custom Vision model is trained in Azure and not locally, we need to upload some images. The data we’ll use comes from the Microsoft Computer Vision Best Practices project. This is a simple set of images containing 4 kinds of objects one might find in a fridge: cans, cartons, milk bottles, and water bottles.
The generic function to add images to a project is
add_images, which takes a vector of filenames, Internet URLs or raw vectors as the images to upload. It returns a vector of image IDs, which are how Custom Vision keeps track of the images it uses.
Let’s upload the fridge objects to the project. The method for classification projects has a
tags argument which can be used to assign labels to the images as they are uploaded. We’ll keep aside 5 images from each class of object to use as validation data.
cans <- dir("fridgeObjects/can", full.names=TRUE) cartons <- dir("fridgeObjects/carton", full.names=TRUE) milk <- dir("fridgeObjects/milk_bottle", full.names=TRUE) water <- dir("fridgeObjects/water_bottle", full.names=TRUE) # upload all but 5 images from cans and cartons, and tag them can_ids <- add_images(testproj, cans[-(1:5)], tags="can") carton_ids <- add_images(testproj, cartons[-(1:5)], tags="carton")
If you don’t tag the images at upload time, you can do so later with
Other image functions to be aware of include
add_image_regions (which is for object detection projects). A useful one is
browse_images, which takes a vector of IDs and displays the corresponding images in your browser.
Training the model
Having uploaded the data, we can train the Custom Vision model with
train_model. This trains the model on the server and returns a model iteration, which is the result of running the training algorithm on the current set of images. Each time you call
train_model, for example to update the model after adding or removing images, you will obtain a different model iteration. In general, you can rely on AzureVision to keep track of the iterations for you, and automatically return the relevant results for the latest iteration.
We can examine the model performance on the training data with the
summary method. For this toy problem, the model manages to obtain a perfect fit.
Obtaining predictions from the trained model is done with the
predict method. By default, this returns the predicted tag (class label) for the image, but you can also get the predicted class probabilities by specifying
## validation_tags ## predicted_tags can carton milk_bottle water_bottle ## can 4 0 0 0 ## carton 0 5 0 0 ## milk_bottle 1 0 5 0 ## water_bottle 0 0 0 5
This shows that the model got 19 out of 20 predictions correct on the validation data, misclassifying one of the cans as a milk bottle.
Publishing to a prediction resource
The code above demonstrates using the training endpoint to obtain predictions, which is really meant only for model testing and validation. In a production setting, we would normally publish a trained model to a Custom Vision prediction resource. Among other things, a user with access to the training endpoint has complete freedom to modify the model and the data, whereas access to the prediction endpoint only allows getting predictions.
Publishing a model requires knowing the Azure resource ID of the prediction resource. Here, we’ll use the resource object that we created earlier; you can also obtain this information from the Azure Portal.
Once a model has been published, we can obtain predictions from the prediction endpoint in a manner very similar to previously. We create a predictive service object with
classification_service, and then call the
predict method. Note that a required input is the project ID; you can supply this directly or via the project object. It may also take some time before a published model shows up on the prediction endpoint.
Sys.sleep(60) # wait for Azure to finish publishing pred_url <- pred_res$properties$endpoint pred_key <- pred_res$list_keys() pred_endp <- customvision_prediction_endpoint(url=pred_url, key=pred_key) project_id <- testproj$project$id pred_svc <- classification_service(pred_endp, project_id, "iteration1") # predictions from prediction endpoint -- same as before predsvc_tags <- predict(pred_svc, validation_imgs) table(predsvc_tags, validation_tags)
## validation_tags ## predsvc_tags can carton milk_bottle water_bottle ## can 4 0 0 0 ## carton 0 5 0 0 ## milk_bottle 1 0 5 0 ## water_bottle 0 0 0 5
Exporting as standalone
As an alternative to deploying the model to an online predictive service resource, for example if you want to create a custom deployment solution, you can also export the model as a standalone object. This is only possible if the project was created to support exporting. The formats supported include:
- ONNX 1.2
- TensorFlow or TensorFlow Lite
- A Docker image for either the Linux, Windows or Raspberry Pi environment
- Vision AI Development Kit (VAIDK)
To export the model, simply call
export_model and specify the target format. This will download the model to your local machine.
AzureVision is part of the AzureR family of packages. This provides a range of tools to facilitate access to Azure services for data scientists working in R, such as AAD authentication, blob and file storage, Resource Manager, container services, Data Explorer (Kusto), and more.
If you are interested in Custom Vision, you may also want to check out CustomVision.ai, which is an interactive frontend for building Custom Vision models.