Image Classification: With Google Cloud AutoML Vision API

1 Tasks:

• The Objectives is to teach how to use Google Cloud AutoML Vision API for images Classification
  • how to upload a labeled dataset to Google Cloud Storage
  • how to connect it to AutoML Vision with a CSV label file.
  • how to train a model with AutoML Vision and evaluate its accuracy.
  • how to generate predictions on your trained model.
• Of Course,please note that running the Google Cloud AutoML Vision API have to pay fee and charges.

1.1 Google Cloud Platform Console

• first to sign in Google Cloud Platform: https://cloud.google.com/

1.2 Set up AutoML Vision

1.3 Cloud Shell

• paste these commands to create environment variables for your Project ID
• Qwiklabs Username, replacing with the user name you logged into the lab with:

export PROJECT_ID=$DEVSHELL_PROJECT_ID
export QWIKLABS_USERNAME=<QWIKLABS_USERNAME>

1.4 Create a Bucket for testing

gsutil mb -p $PROJECT_ID \
    -c regional    \
    -l us-central1 \
    gs://$PROJECT_ID-vcm/

1.5 Navigate to the AutoML UI

1.6 Specify a Google Cloud project

1.7 Click the setup-now button

1.8 Upload training images

• In the GCP console, open the Navigation menu and select Storage > Browser:

• in Cloud Shell, naming YOUR_BUCKET_NAME

export BUCKET=YOUR_BUCKET_NAME

• gsutil command line utility for Cloud Storage to copy the training images into your bucket

gsutil -m cp -r gs://automl-codelab-clouds/* gs://${BUCKET}

• click the Refresh button at the top of the Cloud Storage browser

1.9 Create a dataset

The training data is in Cloud Storage, you need a way for AutoML Vision to access it

• create a CSV file where each row contains a URL to a training image and the associated label for that image.

• run the following command to copy the file to your Cloud Shell instance:

gsutil cp gs://automl-codelab-metadata/data.csv .

• update the CSV with the files in your project:

sed -i -e "s/placeholder/${BUCKET}/g" ./data.csv

• upload this CSV file to your Cloud Storage bucket:

gsutil cp ./data.csv gs://${BUCKET}

• click the Refresh bucket button. Confirm that you see the data.csv file in your bucket

• Navigate back to the AutoML Vision Datasets page

• At the top of the console, click + NEW DATASET.
• Type “clouds” for the Dataset name.
• Leave “Single-label Classification” checked.
• Click CREATE DATASET to continue

• Choose Select a CSV file on Cloud Storage
• add the file name to the URL for the file you just uploaded
• gs://your-project-name-vcm/data.csv.

1.10 Inspect images

1.11 Train your model

1.12 Evaluate your model

1.13 Check Confusion matrix

1.14 Generate predictions