CV Tasks

机器学习

Classification

Input: Image
Output: Class label
Evaluation metric: Accuracy

Localization

Input: Image
Output: Box in the image $(x, y, w, h)$
Evaluation metric: Intersection over union

Object detection

Detection as regression? Need variable sized outputs.
Detection as classification? Need to test many positions and scales, only look at a tiny subset of possible positions (Region proposal: Selective search).