I'm developing a computer vision system, but there are several components that relate more to the NLP field than to CV. I can work with many CV technologies well, but I'm not a specialist in NLP. As I have understood from tutorials, traditional ML algorithms and recurrent neural networks are the main tools used for text processing. But is there something else except what I mentioned?
Yes, of course, there are a lot of different NLP-specific tools, approaches, algorithms, etc. It is true not only for the model-building stage but also for text preprocessing. Learn about different methods of text vectorization, for example, TF-IDF, Bag of words, word2vec, and others. Learn about word stemming, lemmatization, POS-tagging, text cleaning and many many other things.
Actually, NLP (like CV) become more and more independent from general machine learning. It is such a huge field, that many companies require independent positions for NLP, for CV, and for general ML engineers.
Can I use some of my CV knowledge to become more successful in NLP?
CV and NLP are both based on machine learning. So, many core principles are the same. You will see this when you will start to dive deeper into NLP. Neural networks are used in NLP as well as in CV. You are right that RNN is more popular in NLP, but it is also possible to use convolutional neural networks also.
It is very interesting. Can you tell more about CNN in text processing, please?
In fact, there was a similar question earlier on this forum. Here is it: https://imaginghub.com/forum/posts/1304-how-cnn-can-be-used-for-nlp-problems .
Thanks a lot!
Just drop us an email to ...
Show more 
Log in with Google