期刊
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
卷 44, 期 11, 页码 7940-7954出版社
IEEE COMPUTER SOC
DOI: 10.1109/TPAMI.2021.3114089
关键词
Semantics; Training; Tagging; Image retrieval; Correlation; Training data; Binary codes; Image retrieval; multimedia retrieval; multi-modal learning; weakly-supervised learning
资金
- National Natural Science Foundation of China [62133012, 61936006, 62103314, 61876144, 61876145, 62073255]
- Key Research and Development Program of Shaanxi Program [2020ZDLGY04-07]
- Innovation Capability Support Program of Shaanxi Program [2021TD-05]
The paper focuses on using user-tagged images to learn proper hashing functions for image retrieval, proposing a novel weakly-supervised deep hashing framework. By training in two stages, it effectively leverages tagging information and image content for hashing learning.
We are concerned with using user-tagged images to learn proper hashing functions for image retrieval. The benefits are two-fold: (1) we could obtain abundant training data for deep hashing models; (2) tagging data possesses richer semantic information which could help better characterize similarity relationships between images. However, tagging data suffers from noises, vagueness and incompleteness. Different from previous unsupervised or supervised hashing learning, we propose a novel weakly-supervised deep hashing framework which consists of two stages: weakly-supervised pre-training and supervised fine-tuning. The second stage is as usual. In the first stage, we propose two formulations Tag-basEd weakLy-supErvised Modally COoperative hashing Network (TelecomNet) and Generalized TelecomNet (GTelecomNet). Rather than performing supervision on tags, TelecomNet first learns an observed semantic embedding vector for each image from attached tags and then uses it to guide hashing learning. GTelecomNet introduces a novel semantic network to exploit more precise semantic information. By carefully designing the optimization problem, they can well leverage tagging information and image content for hashing learning. The framework is general and does not depend on specific deep hashing methods. Empirical results on real world datasets show that they significantly increase the performance of state-of-the-art deep hashing methods.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据