Application	Illustration	Available Models
Image Classification: recognize an object in an image.		50+ models, including ResNet, MobileNet, DenseNet, VGG, ...
Object Detection: detect multiple objects with their bounding boxes in an image.		Faster RCNN, SSD, Yolo-v3
Semantic Segmentation: associate each pixel of an image with a categorical label.		FCN, PSP, DeepLab v3
Instance Segmentation: associate each pixel of an image with an instance label.		Mask RCNN
Pose Estimation: detect human pose from images.		Simple Pose