zdcs的专栏

因为要预研VQA项目参考，趁GPU满负荷的时间，记录下这个数据集相关笔记：官方网站定义为：Visual Genome 是一个数据集，知识库，不断努力把结构化的图像概念和语言连接起来。使用了众包的方式实现，由李飞飞一位同事 Michael Bernstein 提出。截至今天2016/12/08包含：108077张图片540 万对区域的描述（Region

因为要预研VQA项目参考，趁GPU满负荷的时间，记录下这个数据集相关笔记：官方网站http://www.visualqa.org/目前发布了v1.0, 包含真实图像（MSCOCO 数据集）：204,721 MSCOCO images (all of current train/val/test)614,163 questions6,141,630 groun

Ask, Attend and Answer: Exploring Question-Guided Spatial Attention for Visual Question AnsweringHuijuan Xu and Kate SaenkoDepartment of Computer Science, UMass Lowell, USA hxu1@cs.uml.edu, saenko

Aligning where to see and what to tell: image caption with region-based attention and scene factorizationrXiv:1506.06272v1  [cs.CV]  20 Jun 2015摘要部分:本文提出一种图像文字标注系统利用了图像与句子之间的平行结构在该模型中，

HADAMARD PRODUCT FOR LOW-RANK BILINEAR POOLINGJin-HwaKim Interdisciplinary Program in Cognitive Science Seoul National University Seoul, 08826, Republic of Korea jhkim@bi.snu.ac.krKyoung-WoonOn Sc

Multimodal Compact Bilinear Pooling for Visual Question Answering and Visual GroundingAkiraFukui*1,2 DongHukPark*1 DaylenYang*1 AnnaRohrbach*1,3 TrevorDarrell1 MarcusRohrbach1 1UC Berkeley EECS, CA,

Hierarchical Question-Image Co-Attention for Visual Question AnsweringJiasenLu∗,JianweiYang∗,DhruvBatra∗† ,DeviParikh∗† ∗Virginia Tech,†Georgia Institute of Technology {jiasenlu, jw2yang, dbatra, pa