What is it about?
This paper is proposed to handle the burstiness issue in retrieving landmark images in popular Bag-of-Words based image retrieval baseline with spatial verification. The proposed approach focuses on handling this issue by using the visual word and geometric information, which can be easily embedded into current retrieval pipeline.
Featured Image
Why is it important?
Handling the burstiness issue is critical to improve retrieval accuracy for landmark image retrieval. The burstiness issue will assign over-estimated similarity score for irrelevant images, resulting in a bad retrieval accuracy.
Perspectives
The burstiness issue is often tackled in feature matching process, and we hope this paper can be a small step to tackle the problem after spatial verification. There should be more attention to project in this field.
Mao Wang
National University of Defense Technology
Read the Original
This page is a summary of: Boosting Landmark Retrieval Baseline with Burstiness Detection , IET Computer Vision, November 2017, the Institution of Engineering and Technology (the IET),
DOI: 10.1049/iet-cvi.2016.0504.
You can read the full text:
Contributors
The following have contributed to this page
