Deep Image Retrieval: Learning global representations for image search
Albert Gordo Jon Almazán Jerome Revaud Diane Larlus
Abstract
We propose a novel approach for instance-level image retrieval. It produces a global and compact fixed-length representation for each image by aggregating many region-wise descriptors. In contrast to previous works employing pre-trained deep networks as a black box to produce features, our method leverages a deep architecture trained for the specific task of image retrieval. Our contribution is twofold: (i) we leverage a ranking framework to learn convolution and projection weights that are used to build the region features; and (ii) we employ a region proposal network to learn which regions should be pooled to form the final global descriptor. We show that using clean training data is key to the success of our approach. To that aim, we use a large scale but noisy landmark dataset and develop an automatic cleaning approach. The proposed architecture produces a global image representation in a single forward pass. Our approach significantly outperforms previous approaches based on global descriptors on standard datasets. It even surpasses most prior works based on costly local descriptor indexing and spatial verification111Additional material available at www.xrce.xerox.com/Deep-Ima
中文速览
图像检索领域长期被依赖局部特征匹配加空间验证的传统方法所统治，深度学习方法始终难以追平其精度，根本原因在于没有针对"找同一个物体"这一任务专门训练特征表示。本文提出了一套端到端可训练的图像检索架构：以区域最大激活聚合描述子（R-MAC）为基础，用三路孪生网络（Siamese network）和三元组排序损失（triplet ranking loss）直接优化检索目标，同时引入区域候选网络（Region Proposal Network）替代原有的固定网格采样，让模型自动学习"该聚合哪些区域"；此外还对大规模但噪声严重的地标图像数据集进行了自动清洗，并以清洗后的数据作为训练集，显著提升了学习效果。在五个标准检索数据集上，该方法仅靠单次前向传播生成的全局向量，就超越了所有此前基于全局描述子的方案，甚至在多个基准上超过了需要关键点匹配与空间验证的复杂流程。这项工作证明了"为检索任务量身定制的端到端学习"是深度方法在实例级图像检索中取得突破的关键，为后续大规模视觉检索系统的设计提供了重要参考。
原文 arXiv:1604.01325;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1604.01325v2