Analyze Face Recognition Results

in this step, you will use Kibana in Amazon Elasticsearch Service to visualize the results of facial recognition in real time.

Setting up index pattern

First, open your Kibana dashboard.

After opening Kibana dashboard, you need to set the index pattern at first.

  • Click on the gear icon on the left menu
  • Select Index Patterns on the left, and click on Create index pattern on the right
  • Step 1 of 2: Define index pattern
    • Index pattern: record and click on Next step
  • Step 2 of 2: Configure settings
    • Time Filter field name: timestamp
    • Click on Create index pattern
  • You will see the following screen and confirm that the record types are being analyzed

Set up index patterns

Checking records

Next, check the records registered in Amazon Elasticsearch Service in Kibana.

  • Click the compass icon on the left menu
  • You can check the name and confidence at each time on the screen as follows


Visualizing the results on Kibana

Finally, let’s visualize the recognition results as a graph.

  • Click on the chart icon on the left, then click Create new visualization in the middle of the screen
  • Click on the Vertical Bar
  • In New Data Table / Choose a source, click on record*
  • In the Metrics section on the left, set the following
    • Repeat the following process for each person in your collection
      • Click Metric > + Add and select Y-Axis
      • Aggregation: Average Bucket
      • bucket
        • Aggregation: Filters
        • Filter 1: name.keyword == the name of person in the collection (e.g. name.keyword == alice)
      • Metric
        • Aggregation: Count
        • Custom label: Your name (e.g. alice)
  • Click Buckets > + Add and select X-Axis
  • Configure the following settings
    • Aggregation: Date histogram
  • Click the play button above the Metrics to apply the changes

The following graph will be displayed. This visualizes how many times each person is recognized at each time.


Finally, click on Save at the top of the screen to save the visualization.

Great work, and that’s all the steps for this workshop!

If you want to delete the resources you created in the workshop, proceed to Lab-3.