Computer Vision on AWS

3

Creating Custom Models with Amazon Rekognition Custom Labels

In the last chapter, we learned about Amazon Rekognition’s capability to identify objects in images and videos using built-in APIs. However, there are situations where you’re looking for certain things that have specific meaning to you or your business. For example, imagine you’re an auto parts manufacturer and want to identify different parts such as a crankshaft, torque converter, or radiator. Typically, generic machine learning (ML) models would identify these parts as auto parts, but that label may not be specific enough for your needs. Take another scenario—if you’re a car enthusiast, you’d know the logos of different car manufacturers. If you would like to detect the logos ...