Eliseo  Wolf

Eliseo Wolf

1600322280

Make Animations Feel Pro (with Sarah Drasner) — Learn With Jason

When it comes to animation, there are a few tips & tricks that take them from “neat” to “HOLY WOW” animation master Sarah Drasner teaches us how to take our animations to the next level

#developer #programming

What is GEEK

Buddha Community

Make Animations Feel Pro (with Sarah Drasner) — Learn With Jason

Lets make some Anime using Deep Learning

The motivation for this project was to see how far technology has come in just a few years in the NLP domain especially when it comes to generating creative content. I have explored two text generation techniques by generating Anime synopsis, first with LSTM units which is a relatively old technique and then with a fine tuned GPT2 transformer.

In this post you will see how AI went from creating this piece of nonsense…

A young woman capable : a neuroi laborer of the human , where one are sent back home ? after defeating everything being their resolve the school who knows if all them make about their abilities . however of those called her past student tar barges together when their mysterious high artist are taken up as planned while to eat to fight !

to this piece of art.

A young woman named Haruka is a high school student who has a crush on a mysterious girl named Miki. She is the only one who can remember the name of the girl, and she is determined to find out who she really is.

To get the most out of this post you must have knowledge of :

  • Python programming
  • Pytorch
  • Working of RNNs
  • Transformers

Alright then, lets see some code!


Data Description

The data used here has been scraped from myanimelist, it initially contained over 16000 data points and it was a really messy dataset. I have taken the following steps to clean it:

  • Removed all the weird genres of Anime (if you’re an Anime fan you will know what I`m talking about).
  • Every synopsis contained its source in the end of the description (Eg: Source: myanimelist, Source: crunchyroll etc.) so I have removed that as well.
  • Animes that are based on video games, spin-offs or some adaptation had very small summaries so I removed all the synopses with words less than 30 & I also removed all the synopses which contained the words “spin-off”, “based on”, “music video”, “adaptation”. The logic behind this was that these types of Animes won’t really make our model creative.
  • I have also removed Animes with synopsis words more than 300. This is just to make the training easier (check GPT2 section for more details).
  • Removed symbols.
  • Some descriptions also contained japanese characters so those were also removed.

The following functions take care of all this

import re

	def remove_source(text):
	    cln_text = text
	    if '(Source' in cln_text:
	        cln_text,_,_ = cln_text.partition('(Source')
	    elif '[Written ' in cln_text:
	        cln_text,_,_ = cln_text.partition('[Written')

	    return cln_text

	def clean_synopsis(data):
	    # removing hentai and kids tags
	    data = data[(data.Hentai != 1) & (data.Kids != 1)]
	    synopsis = data.synopsis

	    # removing very small synopsis
	    synopsis = synopsis.apply(lambda x: x if ((len(str(x).strip().split())<=300) and len(str(x).strip().split())>30  ) else -1)
	    synopsis = synopsis[synopsis!=-1]

	    # removing source text
	    synopsis = synopsis.apply(lambda x: remove_source(x))

	    # removing japanese characters
	    synopsis = synopsis.apply(lambda x: re.sub("([^\x00-\x7F])+"," ",x))

	    # remove symbols
	    rx = re.compile('[&#/@`)(;<=\'"$%>]')
	    synopsis = synopsis.apply(lambda x: rx.sub('',x))
	    synopsis = synopsis.apply(lambda x: x.replace('>',""))
	    synopsis = synopsis.apply(lambda x: x.replace('`',""))
	    synopsis = synopsis.apply(lambda x: x.replace(')',""))
	    synopsis = synopsis.apply(lambda x: x.replace('(',""))

	    # removing adaptation animes (some relevant might get deleted but there aren`t a lot so we wont be affected as much)
	    synopsis = synopsis[synopsis.apply(lambda x: 'adaptation' not in str(x).lower())]    
	    synopsis = synopsis[synopsis.apply(lambda x: 'music video' not in str(x).lower())]
	    synopsis = synopsis[synopsis.apply(lambda x: 'based on' not in str(x).lower())]
	    synopsis = synopsis[synopsis.apply(lambda x: 'spin-off' not in str(x).lower())]

	    return synopsis.reset_index(drop=True)
view raw
Clean Synopsis.py hosted with ❤ by GitHub

The LSTM way

The traditional approach for text generation uses recurrent LSTM units. LSTM (or long short term memory) are specifically designed to capture long term dependencies in sequential data which the normal RNNs can’t and it does so by using multiple gates which govern the information that passes from one time step to another.

Intuitively, in a time step the information that reaches an LSTM unit goes through these gates and they decide if the information needs to be _updated, _if they are updated then the old information is forgotten and then this new updated values are sent to the next time step. For a more detailed understanding of LSTMs I would suggest you to go through this blog.

#machine-learning #anime #data-science #deep-learning #deep learning

James Daneil

1580992154

Learn Character Design in After Effects 2D Animation Course

With the help of this course, you can learn to create and animate characters who express with body language in After Effects. Our personal purpose is to help anyone interested in Animation to start practicing with little projects, simple Characters, and most of all, explore the expressiveness of their Body Language and Character Acting. Many people seldom to start learning 2D animation because they are convinced that you need to know how to draw. While drawing skills can help you to improve, that is not the essential skill to do animation. For animation you need to understand the most basic principles in animation, like timing, anticipation, pose to pose. This course is divided into 3 parts theory, rigging and animation which will help you learn how to design characters, character animation and body language expressions. Enroll now and Learn to create 2D Animation in After Effects.

#2d animation #character animation #character rigging #learn animation #animation courses

Jerad  Bailey

Jerad Bailey

1598891580

Google Reveals "What is being Transferred” in Transfer Learning

Recently, researchers from Google proposed the solution of a very fundamental question in the machine learning community — What is being transferred in Transfer Learning? They explained various tools and analyses to address the fundamental question.

The ability to transfer the domain knowledge of one machine in which it is trained on to another where the data is usually scarce is one of the desired capabilities for machines. Researchers around the globe have been using transfer learning in various deep learning applications, including object detection, image classification, medical imaging tasks, among others.

#developers corner #learn transfer learning #machine learning #transfer learning #transfer learning methods #transfer learning resources

Eliseo  Wolf

Eliseo Wolf

1598853480

Make Animations Feel Pro (with Sarah Drasner) — Learn With Jason

When it comes to animation, there are a few tips & tricks that take them from “neat” to “HOLY WOW” animation master Sarah Drasner teaches us how to take our animations to the next level

#developer #programming

Eliseo  Wolf

Eliseo Wolf

1600322280

Make Animations Feel Pro (with Sarah Drasner) — Learn With Jason

When it comes to animation, there are a few tips & tricks that take them from “neat” to “HOLY WOW” animation master Sarah Drasner teaches us how to take our animations to the next level

#developer #programming