Geometric VLAD for Large Scale Image Search

TL;DR

This paper introduces Geometric VLAD, a compact image descriptor that enhances large-scale image search by integrating weak geometric cues, leading to over 15% improvement in retrieval accuracy.

Contribution

The paper proposes Geometric VLAD (gVLAD), a novel extension of VLAD that incorporates keypoint angle information and a new codebook adaptation strategy for improved image retrieval.

Findings

Achieves more than 15% improvement in mAP over benchmarks.

Incorporates weak geometry cues into VLAD framework.

Introduces a clustering-based method for learning membership functions.

Abstract

We present a novel compact image descriptor for large scale image search. Our proposed descriptor - Geometric VLAD (gVLAD) is an extension of VLAD (Vector of Locally Aggregated Descriptors) that incorporates weak geometry information into the VLAD framework. The proposed geometry cues are derived as a membership function over keypoint angles which contain evident and informative information but yet often discarded. A principled technique for learning the membership function by clustering angles is also presented. Further, to address the overhead of iterative codebook training over real-time datasets, a novel codebook adaptation strategy is outlined. Finally, we demonstrate the efficacy of proposed gVLAD based retrieval framework where we achieve more than 15% improvement in mAP over existing benchmarks.