Weakly- and Semi-Supervised Object Detection with Expectation-Maximization Algorithm

Abstract

Object detection when provided image-level labels instead of instance-level labels (i.e., bounding boxes) during training is an important problem in computer vision, since large scale image datasets with instance-level labels are extremely costly to obtain. In this paper, we address this challenging problem by developing an Expectation-Maximization (EM) based object detection method using deep convolutional neural networks (CNNs). Our method is applicable to both the weakly-supervised and semisupervised settings. Extensive experiments on PASCAL VOC 2007 benchmark show that (1) in the weakly supervised setting, our method provides significant detection performance improvement over current state-of-the-art methods, (2) having access to a small number of strongly (instance-level) annotated images, our method can almost match the performace of the fully supervised Fast RCNN.

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

“…Semi-supervised learning. There are several works that assume a few images are annotated with object bounding boxes and the rest still have image-level labels as in WSOD [50,18,19,57]. These are often called semisupervised [50,57].…”

“…In terms of reducing the amount of supervision, the most common setting is weakly-supervised object detection (WSOD) [12,4,40,48,57]. In this setting, we are given a set of images known to contain instances of a certain class as specified by labels, but we do not know the object lo- cations in the form of bounding boxes or otherwise.…”

“…There are mixed approaches where a small number of images are annotated with bounding boxes and labels, and a large amount of images have only image-level labels [50,18,19,57]. Many are related to transfer learning where some classes come with strong supervision (e.g.…”

“…imagelevel labels). This is often referred as a semi-supervised setting [50,57], but there is no consensus.…”

One-Click Annotation with Guided Hierarchical Object Detection

“…Zhang et al [27] has proposed a new state-of-the-art weakly supervised model which combines saliency detection and weakly supervised object detection based on self-paced curriculum learning. There has also been research work where models have been created which can work in both weakly -supervised and semi-supervised paradigm such as [26].…”