Sequential Optimization for Efficient High-Quality Object Proposal Generation

TL;DR

BING++ is a fast and efficient object proposal method that improves localization accuracy over BING by incorporating edges and segments, achieving high-quality proposals with reduced computational cost.

Contribution

The paper introduces BING++, a novel object proposal algorithm that enhances localization accuracy while maintaining high efficiency, using a probabilistic framework and sequential updates.

Findings

BING++ improves localization quality by 18.5% on VOC2007.

BING++ runs at half the speed of BING on CPU.

BING++ achieves comparable performance to state-of-the-art methods with faster runtime.

Abstract

We are motivated by the need for a generic object proposal generation algorithm which achieves good balance between object detection recall, proposal localization quality and computational efficiency. We propose a novel object proposal algorithm, BING++, which inherits the virtue of good computational efficiency of BING but significantly improves its proposal localization quality. At high level we formulate the problem of object proposal generation from a novel probabilistic perspective, based on which our BING++ manages to improve the localization quality by employing edges and segments to estimate object boundaries and update the proposals sequentially. We propose learning the parameters efficiently by searching for approximate solutions in a quantized parameter space for complexity reduction. We demonstrate the generalization of BING++ with the same fixed parameters across different object classes and datasets. Empirically our BING++ can run at half speed of BING on CPU, but significantly improve the localization quality by 18.5% and 16.7% on both VOC2007 and Microhsoft COCO datasets, respectively. Compared with other state-of-the-art approaches, BING++ can achieve comparable performance, but run significantly faster.