Details about the detection part of AWS Rekognition

Question

I'm doing some investigations about AWS Rekognition. There are two issues I need to know but didn't get the answers.

1) how to get the category list of the object detection part.

2) how long does it take to process an image to get object labels in it without considering the data transmission time.

Is there anyone has any ideas?

Answer 1

Using the AWS Command-Line Interface (CLI):

$ time aws rekognition detect-labels \
--image '{"S3Object":{"Bucket":"my-bucket","Name":"fluffy.jpg"}}' \
--region us-east-1
{
    "Labels": [
        {
            "Confidence": 99.2266845703125, 
            "Name": "Plant"
        }, 
        {
            "Confidence": 99.2266845703125, 
            "Name": "Potted Plant"
        }, 
        {
            "Confidence": 92.96235656738281, 
            "Name": "Animal"
        }, 
        {
            "Confidence": 92.96235656738281, 
            "Name": "Bunny"
        }, 
        {
            "Confidence": 92.96235656738281, 
            "Name": "Mammal"
        }, 
        {
            "Confidence": 92.96235656738281, 
            "Name": "Rabbit"
        }, 
        {
            "Confidence": 92.96235656738281, 
            "Name": "Rodent"
        }
    ]
}

real    0m1.646s
user    0m0.264s
sys 0m0.036s