Knowledge Commons of Institute of Automation,CAS
Context-Aware Attention Network for Image-Text Retrieval | |
Zhang, Qi1,2; Lei, Zhen1,2; Zhang, Zhaoxiang1,2; Stan Z. Li3 | |
2020-06-14 | |
会议名称 | IEEE Conference on Computer Vision and Pattern Recognition |
会议日期 | 2020-6-14 |
会议地点 | Seattle, Washington, USA |
摘要 | As a typical cross-modal problem, image-text bidirectional retrieval relies heavily on the joint embedding learning and similarity measure for each image-text pair. It remains challenging because prior works seldom explore semantic correspondences between modalities and semantic correlations in a single modality at the same time. In this work, we propose a unified Context-Aware Attention Network (CAAN), which selectively focuses on critical local fragments (regions and words) by aggregating the global context. Specifically, it simultaneously utilizes global intermodal alignments and intra-modal correlations to discover latent semantic relations. Considering the interactions between images and sentences in the retrieval process, intramodal correlations are derived from the second-order attention of region-word alignments instead of intuitively comparing the distance between original features. Our method achieves fairly competitive results on two generic image-text retrieval datasets Flickr30K and MS-COCO. |
收录类别 | EI |
七大方向——子方向分类 | 图像视频处理与分析 |
文献类型 | 会议论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/39252 |
专题 | 多模态人工智能系统全国重点实验室_生物识别与安全技术 |
通讯作者 | Lei, Zhen |
作者单位 | 1.NLPR, Institute of Automation, Chinese Academy of Sciences, Beijing, China 2.Center for AI Research and Innovation, Westlake University, Hangzhou, China 3.School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing, China |
第一作者单位 | 模式识别国家重点实验室 |
通讯作者单位 | 模式识别国家重点实验室 |
推荐引用方式 GB/T 7714 | Zhang, Qi,Lei, Zhen,Zhang, Zhaoxiang,et al. Context-Aware Attention Network for Image-Text Retrieval[C],2020. |
条目包含的文件 | 下载所有文件 | |||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
PID6410551.pdf(3229KB) | 会议论文 | 开放获取 | CC BY-NC-SA | 浏览 下载 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论