[PAPER-RECORD]

PAPER RECODED

[NO.1] 【第一次找论文作者要源码】

笑死自己了，在写信的时候本来想开头空格调整好，无意中多按了几次回车，导致最终没有调整好就把邮件发了出去！可能被大佬见笑了！还好是基本信息都有表达清楚！下次发重要邮件的时候一定不能把收件人的邮箱先填，不然无脑按触发快捷键就悲剧了，现在就希望作者能回复我的消息le…

[NO.2] 【my blog】

个人博客网站:https://jucway.github.io/

[NO.3]【余弦相似度】

余弦相似度

[NO.4]【Ablation Study】

An ablation study typically refers to removing some “feature” of the model or algorithm, and seeing how that affects performance.

Examples:

An LSTM has 4 gates: feature, input, output, forget. We might ask: are all 4 necessary? What if I remove one? Indeed, lots of experimentation has gone into LSTM variants, the GRU being a notable example (which is simpler).
If certain tricks are used to get an algorithm to work, it’s useful to know whether the algorithm is robust to removing these tricks. For example, DeepMind’s original DQN paper reports using (1) only periodically updating the reference network and (2) using a replay buffer rather than updating online. It’s very useful for the research community to know that both these tricks are necessary, in order to build on top of these results.
If an algorithm is a modification of a previous work, and has multiple differences, researchers want to know what the key difference is.
Simpler is better (inductive prior towards simpler model classes). If you can get the same performance with two models, prefer the simpler one

[NO.5] How to wirte a Paper!

[Abrast]
[Introduction]
you must make your paper easy to read. You’ve

[Related work]
[Main idear]

show simple toy examples to let people get the main idear

[Image model]
[Algorithm]
[Experiments]
Gone are this days of ,“We think this is a great idea and we expect it will be very useful in computer vision. See how it works on this meaningless, contrived problem?”
[Discussion]
Future work?

[Conclsions]
Conclusions, or what this opens up, or how this can change how we appreach computer vision problems.

【How】

General writing tips
Keep the reader upper-most in your mind;

Treat the reader as you would a guest in your house;

尽量用简短的单词表达出你需要的东西就可以，下面就是比较这种方式；

精简的语句表达出想要的结果

Glance at title
Skim abstract
Look at the figure
Read every word

the caption shoud tell the reader what to notice about the figure.

【Part 2】 How to write a good review

Summary Of the Contributions

• The Strengths
  Describe the strengths of the paper in detail

• The Weakness
  Describe the weaknesses of the paper in detail. Provide solid arguments and evidence for your claims.

  • It is not okay to say that something has been done before if you do not provide concrete references.

• 总结：

1. 用自己的话把自己的贡献描述清楚
2. 如果是在之前的工作上做的改进，就可说明与之前的工作，这个工作有什么改变和改进，突出自己的差异和比较；

• 优势： 把自己的优势详细的体现出来。
• 不足：

• Summarizing the paper

1. you should be able to descrite the paper in your own words
2. why is the paper intersting to the CVPR audient?

• Strengths:

3. the authors are trying real hard to tell you about these

• Weaknesses

4. Put on your most constructive "I’m an advisor and mentor"hat.

• Suggestion To Authors

5. Provide recommendations to the authors how they can improve their manuscript

Great opportunity to read other work
保持开放的心，阅读其他reviews rebuttal, discuss the paper and update your review.

How to be a good reviewer

before during after three part to plan

A list of Questions

1. Is the problem clearly work cover the relevant approaches
2. 
   How important is sata?

justification of design choices
contribution
self-make baselines


增量的实验是不被允许的

novelty is in the eye of the beholder（创新在善于观察的眼中）

Could this interest someone at CVPR!!!
是否审稿人会对这个感兴趣
you rarely get chance to review such papers. Try not to mess it up!

Wrong
content organization not English

Helpful Guideline for Reviewing
-









sort by importance, and answer precisely concisely, unambiguously