Deep Learning is not only useful for Image Classification (whether an image contains a certain object or not), but a host of other vision tasks as well :
- Semantic Segmentation : Pixel-wise classification to identify the class of objects in an image
- Classification + Localization : Classify and locate a desired object via a bounding box
- Object Detection : Identify multiple objects and draw bounding boxes around them
- Instance Segmentation : Similar to Semantic Segmentation but mark (pixel-wise) multiple instances of the same object separately.
