Foreground Object Search by Distilling Composite Image Feature

TL;DR

This paper introduces DiscoFOS, a novel foreground object search method that distills composite image features from a discriminator, achieving efficient and accurate retrieval, and provides two new datasets for the task.

Contribution

The paper proposes a distillation-based FOS method using a discriminator as teacher and releases two new datasets for the task.

Findings

DiscoFOS outperforms previous methods in retrieval accuracy.

The method significantly reduces computation time compared to discriminator-based approaches.

Extensive experiments validate the effectiveness of the proposed approach.

Abstract

Foreground object search (FOS) aims to find compatible foreground objects for a given background image, producing realistic composite image. We observe that competitive retrieval performance could be achieved by using a discriminator to predict the compatibility of composite image, but this approach has unaffordable time cost. To this end, we propose a novel FOS method via distilling composite feature (DiscoFOS). Specifically, the abovementioned discriminator serves as teacher network. The student network employs two encoders to extract foreground feature and background feature. Their interaction output is enforced to match the composite image feature from the teacher network. Additionally, previous works did not release their datasets, so we contribute two datasets for FOS task: S-FOSD dataset with synthetic composite images and R-FOSD dataset with real composite images. Extensive experiments on our two datasets demonstrate the superiority of the proposed method over previous approaches. The dataset and code are available at https://github.com/bcmi/Foreground-Object-Search-Dataset-FOSD.