过去一段时间在研究cross-modal hashing领域,这个领域可以看作是cross-modal retrieval领域的扩展,也可以看作是image retrieval hashing领域的扩展。该领域在15年通过发表于CVPR上的论文SePH("Semantics-preserving hashing for cross-view retrieval")逐渐受到大家关注,也是从这时开始,国人开始大量引入这个领域。 到2020年...
torchcmh是一个基于PyTorch的深度跨模态hashing库 包含: 数据可视化 17-19年有名的baseline方法 多个数据集读取的API 损失函数API 配置调用 数据集 包括了4个我自己制作的数据集(Mirflickr25k, Nus Wide, MS coco, IAPR TC-12) 如果需要使用可以下载对应的.mat文件并去官网下载对应的数据集 下载方式见数据集目录...
现有的代表性方法包括跨模态相似度敏感哈希(cross modality similarity sensitive hashing,CMSSH,通过最小化不同模态的相似样本之间的汉明距离, 最大化不同模态的不相似样本间的汉明距离, 学习哈希函数),跨视图哈希(cross view hashing, CVH, 把谱哈希扩展到 跨模态检索, 通过最小化加权距离, 保持相似样本 (模态内...
论文-Deep Cross-Modal Hashing Deep Cross-Modal Hashing 关键词:cross-Modal, deep learning 因为cross-modal hashing的低存储和高效的查询能力,被广泛用于多媒体的相似度检索中。本论文提出了一个deep cross-modal hashing(DCMH),首次将feature learning和hash-code learning结合到同一个框架中。 DCMH的框架: 该...
To address this issue, we propose a novel cross-modal hashing, termed as Category Structure Preserving Hashing (CSPH), for cross-modal retrieval. In CSPH, category-specific distribution is preserved by a structure-preserving regularization term during the hash learning. Compared with existing ...
多模态检索Deep Cross-Modal Hashing 技术标签:跨模态检索 什么是多模态检索? 现实生活中常有图搜图,文本搜文本,视频搜视频的应用,这些都是单模态检索。多模态检索就是,不同类别之间的搜索,比如用文本搜图,用图搜文本等,这类情况称为多模态检索。 这篇论文的意义? 传统的都是手动提取特征方法,这篇论文将特征...
Unlike existing cross-modal hashing methods that learn hash functions in the form of numeric quantization of linear projections, the proposed hash learning algorithm encodes features' ranking properties and takes advantage of rank correlations which are known to be scale-invariant, numerically stable ...
跨模态检索Coupled CycleGAN: Unsupervised Hashing Network for Cross-Modal Retrieval 核心思想 本论文是无监督方法,主要由两层循环对抗网络构成,外层的循环对抗网络主要是使不同模态提取更有代表性的公共特征向量,内层循环对抗网络使学的高质量的哈希编码...}GfI−>T(是一个encode->decode过程),生成Ffake...
Recently, cross-modal hashing (CMH) methods have attracted much attention. Many methods have been explored; however, there are still some issues that need to be further considered. 1) How to efficiently construct the correlations among heterogeneous modalities. 2) How to solve the NP-hard optimi...
Cross-modal hashing encodes multimodal data into a common binary space, which can efficiently measure correlations between cross-modal instances. However, most existing cross-modal hashing retrieval methods are difficult to handle the heterogeneity problem between different modalities, and the performance ...