Skip to content


Here are 92 public repositories matching this topic...

pravj commented Oct 2, 2018

Given that the accuracy of an OCR system depends on the following (not limited to) factors:

  • Distribution of text in the image
  • Quality of image

It doesn't result in a correct word detection all the times. For example, a word Shoe on screen might result in Shot for a distorted image.

We can experiment with the spell correction to see if that can help in such situations by using the h

lovit commented Dec 6, 2018

현재 document vector inference 는 학습 데이터에 존재하는 단어만을 이용한다. word vector inference 이후 word representation 을 append 할 때, transformer 에도 이 정보를 함께 append 하면 이후 새로운 단어도 이용되어 doc vec inference 가 가능하다.

doc vec inference 함수에 입력되는 (doc, term) matrix 에 새로운 단어가 포함되어 있을 경우, 이 단어들에 대하여 word vector inference 를 먼저 수행한 뒤, doc vec inference 를 할 수도 있다

Improve this page

Add a description, image, and links to the word-embedding topic page so that developers can more easily learn about it.

Curate this topic

Add this topic to your repo

To associate your repository with the word-embedding topic, visit your repo's landing page and select "manage topics."

Learn more

You can’t perform that action at this time.