Discovering the CV built-in algorithms in Amazon SageMaker
SageMaker includes three CV algorithms, based on proven deep learning networks. In this section, you'll learn about these algorithms, what kind of problems they can help you solve, and what their training scenarios are:
- Image classification assigns one or more labels to an image.
- Object detection detects and classifies objects in an image.
- Semantic segmentation assigns every pixel of an image to a specific class.
Discovering the image classification algorithm
Starting from an input image, the image classification algorithm predicts a probability for each class present in the training dataset. This algorithm is based on the ResNet convolutional neural network (https://arxiv.org/abs/1512.03385). Published in 2015, ResNet won the ILSVRC classification task that same year (http://www.image-net.org/challenges/LSVRC/). Since then, it has become a popular and versatile choice for image classification...