Enhancing Weakly-Supervised Object Detection on Static Images through (Hallucinated) Motion

TL;DR

This paper introduces a novel approach to improve weakly-supervised object detection in static images by hallucinating motion information and integrating it into the learning process, leading to better detection performance.

Contribution

The study proposes a method to hallucinate motion from static images and incorporate it into WSOD, addressing camera motion and selectively training images, which is a new approach in this domain.

Findings

Improved detection accuracy on COCO and YouTube-BB datasets.

Effective hallucination of motion enhances WSOD performance.

Addressing camera motion improves robustness.

Abstract

While motion has garnered attention in various tasks, its potential as a modality for weakly-supervised object detection (WSOD) in static images remains unexplored. Our study introduces an approach to enhance WSOD methods by integrating motion information. This method involves leveraging hallucinated motion from static images to improve WSOD on image datasets, utilizing a Siamese network for enhanced representation learning with motion, addressing camera motion through motion normalization, and selectively training images based on object motion. Experimental validation on the COCO and YouTube-BB datasets demonstrates improvements over a state-of-the-art method.