Weakly Supervised Object Localization with Multi-fold Multiple Instance Learning

TL;DR

This paper introduces a multi-fold multiple instance learning method for weakly supervised object localization, effectively improving accuracy without requiring bounding box annotations by iteratively training and refining object locations.

Contribution

The paper proposes a novel multi-fold MIL procedure and a window refinement method to enhance weakly supervised object localization accuracy, especially with high-dimensional features.

Findings

Effective localization on PASCAL VOC 2007

Prevents premature convergence to incorrect locations

Improves accuracy with objectness prior

Abstract

Object category localization is a challenging problem in computer vision. Standard supervised training requires bounding box annotations of object instances. This time-consuming annotation process is sidestepped in weakly supervised learning. In this case, the supervised information is restricted to binary labels that indicate the absence/presence of object instances in the image, without their locations. We follow a multiple-instance learning approach that iteratively trains the detector and infers the object locations in the positive training images. Our main contribution is a multi-fold multiple instance learning procedure, which prevents training from prematurely locking onto erroneous object locations. This procedure is particularly important when using high-dimensional representations, such as Fisher vectors and convolutional neural network features. We also propose a window refinement method, which improves the localization accuracy by incorporating an objectness prior. We present a detailed experimental evaluation using the PASCAL VOC 2007 dataset, which verifies the effectiveness of our approach.