Using AWS services to perform pet face recognition

Question

I want to build a camera that automatically recognizes a specific pet with image or video recognition. What AWS service can I use to identify an individual pet (not just the pet type). I've tried to use AWS Rekognition, but it can only differentiate between animal types, race, or color. Amazon SageMaker could be another option to create a completely new mode, but is very costly. What other AWS services can I use to identify specific pets?

Accepted Answer

You can use Amazon Rekognition Custom Labels to use single class object detection to identify or classify a specific animal. However, note that Amazon Rekognition Custom Labels do not perform animal face recognition. It only classifies the image or object.

For example, you can train your detection model to identify an animal based on the images you provide for that label. For more information about using Amazon Rekognition Custom Labels, see this blog: https://aws.amazon.com/blogs/machine-learning/training-a-custom-single-class-object-detection-model-with-amazon-rekognition-custom-labels/.

To use frames from a video with Custom Labels, see[ Video analysis ](https://docs.aws.amazon.com/rekognition/latest/customlabels-dg/ex-video-extraction.html).

Using AWS services to perform pet face recognition

相關內容