这是一篇关于跨模态检索(Cross-Modal Retrieval)的paper,在2017的ACM Multimedia上也是拿了Best Paper Award。文章主要利用了Adversarial Learning和Triplet Constraint将Image与Text映射到Common Subspace,这样在Subspace的representations就可以直接进行比较,方便检索等其他操作。模型本身被称为ACMR,第一次接触到Domain Adaption...
Introduction 作者提出了一个新的跨模态检索框架 Adversarial Cross-Model Retrieval (ACMR),其利用对抗学习来缩小不同模态特征的gap。下图为框架图: Proposed Method 问题定义: 每对样本的特征定义为: ,每对样本搭配一个语义标签向量 ,其中 c 为语义类的数量,如果第 i 个样本包含了语义 j,则 。数据包含三个矩阵...
本文探讨跨模态检索（Cross-Modal Retrieval）领域，特别是Adversarial Cross-Modal Retrieval（ACMR）这一具有创新性的方法。ACMR在2017年ACM Multimedia会议上获得最佳论文奖，为跨模态数据的检索提供了新的思路。该方法利用对抗学习（Adversarial Learning）和三元约束（Triplet Constraint）将图像和文本映射到公共...
论文地址: https://www.researchgate.net/publication/320541510_Adversarial_Cross-Modal_Retrievalwww.researchgate.net 来源:ACM Multimedia 2017 作者:电子科技大学英才实验学院2014级本科生王泊锟同学以第一作者身份发表,获ACM Multimedia 2017会议最佳论文奖... 查看原文 【论文解读 KDD 2018 | EANN】Event ...
Cross-modal Hashing has received a lot of attentions in the field of cross-modal retrieval due to its high retrieval efficiency and low storage cost. Most of the existing cross-modal Hashing methods learn Hash codes directly from multimod
Self-Supervised Adversarial Hashing Networks for Cross-Modal Retrieval译文 用于跨模式检索的自监督对抗哈希网络 摘要 由于深入学习的成功,跨模式检索最近取得了显著的进展。然而,仍然存在着一个关键的瓶颈:如何弥合情态差异,进一步提高检索的准确性。在本文中,我们提出了一种自我监督的对抗性散列(SSAH)方法,它是早期...
阅读笔记 Modality-specific and shared generative adversarial network for cross-modal retrieval,程序员大本营,技术文章内容聚合第一站。
阅读笔记 Modality-specific and shared generative adversarial network for cross-modal retrieval 这一篇论文讲的是使用多模态来进行图片的检索, 通过文字检索出最好的图片,模型结构如下: 文章提出两个特征概念 modality-specific 模态独立特征 modality-shared 模态分享特征,也可以理解为共同特征...
摘要: Cross-modal retrieval has become a highlighted research topic, to provide flexible retrieval experience across multimedia data such as image, video, text and audio. The core of existing cross-modal...关键词: Cross-modal retrieval Adversarial learning Metric learning ...
As the rapid growth of multi-modal data, hashing methods for cross-modal retrieval have received considerable attention. Deep-networks-based cross-modal hashing methods are appealing as they can integrate feature learning and hash coding into end-to-end trainable frameworks. However, it is still ch...