상세정보
학술대회
Multi-stage Image Retrieval based on Feature Augmentation with Truncated Polynomial Weight
- 발행일
- 201810
- 출처
- International Conference on Information and Communication Technology Convergence (ICTC) 2018, pp.480-483
- 협약과제
- 18HS3800, 온-오프라인에서의 콘텐츠 비주얼 브라우징 기술 개발, 이승재
- 초록
- In this paper, we propose an effective image retrieval method. Based on a conventional global image representation, multi-stage image retrieval pipeline with feature augmentation is constructed to improve retrieval accuracy. To suppress irrelevant images while boosting relevant images, a novel weighting scheme for feature augmentation is introduced. In addition, the relationship between database images is leveraged to update or re-rank shortlist of the retrieved images. The proposed method was evaluated on Google-landmarks dataset, and the experimental results validate the effectiveness of the proposed method.
- KSP 제안 키워드
- Feature augmentation, Image retrieval, Multi-stage, image representation, re-rank, retrieval accuracy, retrieval method, weighting scheme