Training Object Detectors from Few Weakly-Labeled and Many Unlabeled Images

Weakly-supervised object detection attempts to limit the amount of supervision by dispensing the need for bounding boxes, but still assumes image-level labels on the entire training set are available. In this work, we study the problem of training an object detector from one or few images with image-level labels and a larger set of completely unlabeled images. This is an extreme case of semi-supervised learning where the labeled data are not enough to bootstrap the learning of a detector. Our solution is to train a weakly-supervised student model from image-level pseudo-labels generated on the unlabeled set by a teacher model, bootstrapped by region-level similarities to labeled images. Building upon a recent representative weakly-supervised pipeline PCL, our method shows the capability of effectively making using of more unlabeled images and achieve performance competitive or superior to many state of the art weakly-supervised detection solutions.