Diagnosing State-Of-The-Art Object Proposal Methods

TL;DR

This paper performs a comprehensive meta-analysis of state-of-the-art object proposal methods, examining how object-level characteristics affect their performance and identifying limitations related to object size, view, and appearance.

Contribution

It is the first study to analyze the impact of object-level features on object proposal methods, providing insights for future improvements.

Findings

Existing methods struggle with small objects and non-iconic views.

Performance is affected by color contrast and shape regularity.

Lessons for selecting and designing object proposal techniques are shared.

Abstract

Object proposal has become a popular paradigm to replace exhaustive sliding window search in current top-performing methods in PASCAL VOC and ImageNet. Recently, Hosang et al. conduct the first unified study of existing methods' in terms of various image-level degradations. On the other hand, the vital question "what object-level characteristics really affect existing methods' performance?" is not yet answered. Inspired by Hoiem et al.'s work in categorical object detection, this paper conducts the first meta-analysis of various object-level characteristics' impact on state-of-the-art object proposal methods. Specifically, we examine the effects of object size, aspect ratio, iconic view, color contrast, shape regularity and texture. We also analyse existing methods' localization accuracy and latency for various PASCAL VOC object classes. Our study reveals the limitations of existing methods in terms of non-iconic view, small object size, low color contrast, shape regularity etc. Based on our observations, lessons are also learned and shared with respect to the selection of existing object proposal technologies as well as the design of the future ones.