Random VLAD based Deep Hashing for Efficient Image Retrieval

• Main Authors: Weng, Li, Ye, Lingzhi, Tian, Jiangmin, Cao, Jiuwen, Wang, Jianzhong
• Format: Journal Article
• Language: English
• Published: 06.02.2020
• Subjects: Computer Science - Computer Vision and Pattern Recognition; Computer Science - Information Retrieval
• DOI: 10.48550/arxiv.2002.02333

Image hash algorithms generate compact binary representations that can be quickly matched by Hamming distance, thus become an efficient solution for large-scale image retrieval. This paper proposes RV-SSDH, a deep image hash algorithm that incorporates the classical VLAD (vector of locally aggregated descriptors) architecture into neural networks. Specifically, a novel neural network component is formed by coupling a random VLAD layer with a latent hash layer through a transform layer. This component can be combined with convolutional layers to realize a hash algorithm. We implement RV-SSDH as a point-wise algorithm that can be efficiently trained by minimizing classification error and quantization loss. Comprehensive experiments show this new architecture significantly outperforms baselines such as NetVLAD and SSDH, and offers a cost-effective trade-off in the state-of-the-art. In addition, the proposed random VLAD layer leads to satisfactory accuracy with low complexity, thus shows promising potentials as an alternative to NetVLAD.