A System for Rapid Interactive Training of Object Detectors

Machine learning approaches have become the de-facto standard for creating object detectors (such as face and pedestrian detectors) which are robust to lighting, viewpoint, and pose. Generating sufficiently large labeled data sets to support accurate training is often the most challenging problem. To address this, the active learning paradigm suggests interactive user input, creating an initial classifier based on a few samples and refining that classifier by identifying errors and re-training. In this paper we seek to maximize the efficiency of the user input; minimizing the number of labels the user must provide and minimizing the accuracy with which the user must identify the object. We propose, implement, and test a system that allows an untrained user to create high-quality classifiers in minutes for many different types of objects in arbitrary scenes.