Bridging Saliency Detection to Weakly Supervised Object Detection Based on Self-paced Curriculum Learning

TL;DR

This paper introduces a novel approach that integrates saliency detection with weakly supervised object detection using self-paced curriculum learning, leading to improved detection accuracy in complex scenes.

Contribution

It is one of the first to connect saliency detection with WOD through self-paced curriculum learning, enhancing the learning process from easy to hard examples.

Findings

Achieves state-of-the-art results in weakly supervised object detection.

Effectively guides the learning process using saliency-based priors.

Demonstrates robustness in complex visual scenes.

Abstract

Weakly-supervised object detection (WOD) is a challenging problems in computer vision. The key problem is to simultaneously infer the exact object locations in the training images and train the object detectors, given only the training images with weak image-level labels. Intuitively, by simulating the selective attention mechanism of human visual system, saliency detection technique can select attractive objects in scenes and thus is a potential way to provide useful priors for WOD. However, the way to adopt saliency detection in WOD is not trivial since the detected saliency region might be possibly highly ambiguous in complex cases. To this end, this paper first comprehensively analyzes the challenges in applying saliency detection to WOD. Then, we make one of the earliest efforts to bridge saliency detection to WOD via the self-paced curriculum learning, which can guide the learning procedure to gradually achieve faithful knowledge of multi-class objects from easy to hard. The experimental results demonstrate that the proposed approach can successfully bridge saliency detection and WOD tasks and achieve the state-of-the-art object detection results under the weak supervision.