Link of the Paper: https://arxiv.org/abs/1711.09151

Motivation:

  • LSTM units are complex and inherently sequential across time.
  • Convolutional networks have shown advantages on machine translation and conditional image generation.

Innovation:

  • The authors develop a convolutional ( CNN-based ) image captioning method that shows comparable performance to an LSTM based method on standard metrics.

    

  • The authors analyze the characteristics of CNN and LSTM nets and provide useful insights such as -- CNNs produce more entropy ( useful for diverse predictions ), better classification accuracy, and do not suffer from vanishing gradients.

Improvement:

  • Improved performance with a CNN model that uses Attention Mechanism to leverage spatial image features.

General Points:

  • Image Captioning is applicable to virtual assistants, editing tools, image indexing and support of the disabled.
  • Image Captioning is a basic ingredient for more complex operations such as storytelling and visual summarization.
  • An illustration of a classical RNN architecture for image captioning is provided below.

Paper Reading - Convolutional Image Captioning ( CVPR 2018 )的更多相关文章

  1. Paper Read: Convolutional Image Captioning

    Convolutional Image Captioning 2018-11-04 20:42:07 Paper: http://openaccess.thecvf.com/content_cvpr_ ...

  2. Paper Reading - Learning to Evaluate Image Captioning ( CVPR 2018 ) ★

    Link of the Paper: https://arxiv.org/abs/1806.06422 Innovations: The authors propose a novel learnin ...

  3. Paper Reading - Convolutional Sequence to Sequence Learning ( CoRR 2017 ) ★

    Link of the Paper: https://arxiv.org/abs/1705.03122 Motivation: Compared to recurrent layers, convol ...

  4. Paper Reading: Stereo DSO

    开篇第一篇就写一个paper reading吧,用markdown+vim写东西切换中英文挺麻烦的,有些就偷懒都用英文写了. Stereo DSO: Large-Scale Direct Sparse ...

  5. 爬取CVPR 2018过程中遇到的坑

    爬取 CVPR 2018 过程中遇到的坑 使用语言及模块 语言: Python 3.6.6 模块: re requests lxml bs4 过程 一开始都挺顺利的,先获取到所有文章的链接再逐个爬取获 ...

  6. 在矩池云上复现 CVPR 2018 LearningToCompare_FSL 环境

    这是 CVPR 2018 的一篇少样本学习论文:Learning to Compare: Relation Network for Few-Shot Learning 源码地址:https://git ...

  7. Paper Reading - Long-term Recurrent Convolutional Networks for Visual Recognition and Description ( CVPR 2015 )

    Link of the Paper: https://arxiv.org/abs/1411.4389 Main Points: A novel Recurrent Convolutional Arch ...

  8. Paper Reading - CNN+CNN: Convolutional Decoders for Image Captioning

    Link of the Paper: https://arxiv.org/abs/1805.09019 Innovations: The authors propose a CNN + CNN fra ...

  9. Paper Reading - Deep Captioning with Multimodal Recurrent Neural Networks ( m-RNN ) ( ICLR 2015 ) ★

    Link of the Paper: https://arxiv.org/pdf/1412.6632.pdf Main Points: The authors propose a multimodal ...

随机推荐

  1. Folyd + 路径存储

    一.Folyd 算法原理 如果 AB + AC < BC 那么, BC最短路就要经过 A. 在算法进行过程中,应该是 ,B-A 有很多路径,B 代表这些路径权值之和,A-C也有很多路径,C是这些 ...

  2. chromium之ScopedNSAutoreleasePool浅析

    上代码,看看注释 ScopedNSAutoreleasePool只有Mac系统特有的,也可以理解为OC特有的函数, 其他系统为空实现   // On the Mac, ScopedNSAutorele ...

  3. Linux 学习第四天

    Linux学习第四天 一.常用命令 1.tar  (压缩.解压) A.添加压缩包  tar czvf 压缩包名称.tar.gz 源文件 B.添加压缩包  tar cjvf 压缩包名称.tar.bz2 ...

  4. ES5拓展

    一.JSON拓展 1.JSON.parse(str,fun):将JSON字符串转为js对象 两个参数:str表示要处理的字符串:fun处理函数,函数有两个参数,属性名.属性值 // 定义json字符串 ...

  5. Python学习手册之数据封装、类方法、静态方法和属性函数

    在上一篇文章中,我们介绍了 Python 的内部方法.操作符重载和对象生命周期,现在我们介绍 Python 的数据封装.类方法.静态方法和属性函数.查看上一篇文章请点击:https://www.cnb ...

  6. centos配置NTP服务器

    时间服务器: NTP(Network Time Protocol,网络时间协议)是用来使用网络中的各个计算机时间同步的一种协议,NTP服务器就是利用NTP协议提供时间同步服务的. 一.环境准备: 1. ...

  7. C语言实现 "谁是凶手?"

    日本某地发生了一件谋杀案,警察通过排查确定杀人凶手必为4个嫌疑犯的一个.以下为4个嫌疑犯的供词.A说:不是我.   a=0B说:是C.   c=1 C说:是D.      d=1D说:C在胡说    ...

  8. HttpClient的Content-Type设置

    HttpClient的Content-Type设置 最近在对接公司内容的一个云服务的时候,遇到一个问题,就是如果使用HttpClient如何设置post时候的Content-Type? public ...

  9. java 第七章 面向对象高级特性

    一.类的继承 (一)继承的含义 1.在Java中定义一个类时,让该类通过关键字extends继承一个已有的类,这就是类的继承(泛化). 2.被继承的类称为父类(超类,基类),新的类称为子类(派生类). ...

  10. 北京Uber优步司机奖励政策(12月21日)

    滴快车单单2.5倍,注册地址:http://www.udache.com/ 如何注册Uber司机(全国版最新最详细注册流程)/月入2万/不用抢单:http://www.cnblogs.com/mfry ...