Research @ Mangaki Recommandation d'anime et de mangas

Wanna join? Contact us or say hi on Twitter!


From fiction to reality

De la fiction est née la réalité. Ne l’oublie jamais.
It’s truth that came from fiction. Always remember that.” – Paprika

Here are a few examples.

Melon Cream Soda in Haruhi Suzumiya

In this recorded talk from Yoshiya Makita et al. from Ritsumeikan University at the research symposium of Anime Expo 2017, we learn that:

  • The Haruhi Suzumiya series are based on real-life locations in Nishinomiya (halfway between Kobe and Osaka).
  • Even the café in which Yuki orders a Melon Cream Soda exists in real life (although it has moved from the original location).
  • The Melon Cream Soda was not originally available there, but presumably due to pressure from fans, it is now part of the menu.

That’s the whole point: from fiction it became reality.

See more photos of our Haruhi pilgrimage in Nishinomiya.

Uniforms from Kyoto Kogakuin High School

Back in 2016 when Mangaki received a prize from the Japan Foundation in Paris, the vice-president of Kyoto Manga Museum told us that they were struggling to make manga culture accepted in Japan (not as a subculture). He had such an example of fiction that becomes reality:

Source: Kyoto Kogakuin High School, Wikipedia

  • Kyoto Kogakuin High School wanted to increase their girl rate.
  • They got inspired from a manga called Taihen Yoku Dekimashita (Well Done) and changed their uniforms to match the ones from the manga.
  • Did it worked? I don’t know, but it got covered by many media in Japan.

Sources: Pinterest, Japan Times, Japan Info, Kotaku

If you know other examples, let us know in the comments!

Mangaki on Earth (MoE): visualize anime embeddings

So actually in Mangaki, our algorithms allow us to learn a latent representation (also called embedding) of every anime or manga and every user, so that people like anime in their direction.

Embeddings in Mangaki

So for example, people who like Steins;Gate and Durarara!! are usually not the same than the ones that like Fairy Tail or Naruto.

Using your ratings, we can find where you are in this map, and provide recommendations to you accordingly.

Where are you?

To know more, here are some resources:

And as I was bored, I provided here a t-SNE embedding on France’s map. We called it Mangaki on Earth (MoE).

Mangaki embeddings on France map

If you want to know where you are on the map, feel free to get in touch!

AI for Manga & Anime

At Anime Expo 2018 in Los Angeles, we gave the keynote AI for Manga & Anime (AIMA_AX)!

The main goal of this keynote was to showcase amazing applications of AI to mangas & anime series.

AIMA banner by Jerry Li

Here are some pictures and slides.

Create Anime Characters using AI

Crypko and MakeGirls.Moe were presented by Jingtao Tian (3rd on the image), Yanghua Jin (2nd) and Minjun Li (1st).

Make.Girls.Moe got 1 million views the first 10 days. Our research proceedings were quickly sold out at the Comic Market #92 (Tokyo) in Summer 2017.

Here is an example of what can be achieved using GANs to generate anime characters.

MakeGirlsMoe

About the authors

Using Posters to recommend anime and mangas

Mangaki, a anime/manga recommender system presented by Jill-Jênn Vie from RIKEN AIP, Tokyo.

Everyone regularly ask themselves what movie, series or book they should watch next, according to their taste. Mangaki is a award-winning website that innovates access to Japanese culture through a recommender system. When a user shows up, our algorithm asks them to rate a few works. Based on their answers, they receive a personalized to-watch list of anime & manga, by geometrically positioning their ratings within those collected from other users, and using deep learning to extract information from manga covers or anime posters.

Mangaki gathered 330k ratings from 2,000 people over 11,000 anime & manga works. In Summer 2017, we released them for a data challenge organized by Kyoto University that attracted 31 submissions from 11 countries. Mangaki was awarded the first prize by the Japan Foundation (Paris branch), and an open source award by Microsoft Ventures.

About the author

Automatic Manga Colorization

PaintsChainer

The first part, PaintsChainer, was introduced by Taizan Yonetsuji from Preferred Networks, Japan.

Here is the kind of work that can be achieved using PaintsChainer.

PaintsChainer

style2paints

The second part, style2paints, was presented by LvMing Zhang from Soochow University, China.

Here is the kind of work that can be achieved using style2paints.

style2paints

About the authors

Manga Style Transfer

Cross-Domain Translation of Human Portraits.

Here is another example of what can be achieved using TwinGAN.

Manga style transfer

About the authors

Don’t hesitate to contact the authors to know more! And you can browse the photos of the keynote.

AI for Manga & Anime (AIMA)

We are pleased to give a keynote at the Anime Expo conference in Los Angeles, on July 5!

AI has given rise to AlphaGo and self-driving cars. What about anime? Using deep learning, we can automatically generate the perfect waifu (or husbando) for you, or prioritize your watchlist. Join us for a showcase of amazing research including manga style transfer, automatic colorization, and more!

And here is the current line-up of speakers:

Create Anime Characters using AI

We all love anime characters and are tempted to create our own, but most of us cannot do that because we are not professional artists. AI comes to rescue: on MakeGirlsMoe, you can just specify attributes (such as blonde/twin tailed/smiling) and our deep neural network will generate automatically an anime character at a professional level of quality! Our recent research is targeting style transfer from IRL pictures to manga characters.

Make.Girls.Moe got 1 million views the first 10 days. Our research proceedings were quickly sold out at the Comic Market #92 (Tokyo) in Summer 2017.

MakeGirlsMoe

MakeGirlsMoeTechnical Report (NIPS Workshop for Creativity & Design)
CrypkoWhite paper

Using Posters to recommend anime and mangas

Everyone regularly ask themselves what movie, series or book they should watch next, according to their taste. Mangaki is a award-winning website that innovates access to Japanese culture through a recommender system. When a user shows up, our algorithm asks them to rate a few works. Based on their answers, they receive a personalized to-watch list of anime & manga, by geometrically positioning their ratings within those collected from other users, and using deep learning to extract information from manga covers or anime posters.

Mangaki gathered 330k ratings from 2,000 people over 11,000 anime & manga works. In Summer 2017, we released them for a data challenge organized by Kyoto University that attracted 31 submissions from 11 countries. Mangaki was awarded the first prize by the Japan Foundation (Paris branch), and an open source award by Microsoft Ventures.

Tool: Illustration2Vec by Yusuke Matsui
MangakiPress release – Technical report: the BALSE algorithm

Automatic Manga Colorization

PaintsTransfer

PaintsTransfer + GitHub

PaintsChainer

PaintsChainer by PFN

Manga Style Transfer

Cross-Domain Translation of Human Portraits.

Manga style transfer

Slide from Yanghua Jin’s presentation Creating Anime Characters with GAN at the Tokyo Deep Learning Workshop held in RIKEN AIP on March 21, 2018.

Blog post about TwinGAN

Don’t miss it!

Mangaki Data Challenge Winners

(Cet article est aussi disponible en français.)

The Mangaki Data Challenge

From July 1 to October 1, Mangaki and the Kashima Lab of Kyoto University organized the Mangaki Data Challenge.

The contest was announced at Anime Expo on July 2, 2017, in Los Angeles!

Statement

Participants had to determine whether certain users are interested or not in some anime or manga. Competitors had access to Mangaki ratings as open data.

Read the full problem statement in French, English or Japanese.

Leaderboard

1st prize.

GeniusIke (Microsoft, China), who wins a background artbook Your Name. and the OST of Shaft’s Fireworks movie!

2nd prize.

ηzw, who wins a subscription to the anime streaming website Wakanim.

3rd prize.

karekyasu, who wins 2 collector JoJolion ecocups designed by Sedeto.

See the full leaderboard on University of Big Data.

What are the winning solutions?

The winner, GeniusIke (AUC = 86%), described his solution in a blog post and published his code on GitHub!

Please note that a simple solution that predicts a linear combination of the training set allowed BC to reach the 5th place with 82.6% AUC!
(See the leaderboard and this notebook.)

What is Mangaki’s score?

We got 81% AUC with a gradient boosting tree with which we could have been 8th. See our described solution!

We are wondering, according to our recent research article (BALSE, accepted at MANPU 2017), if using posters would have allowed us to improve this score significantly!

What countries did participate?

  • France: 13
  • Japan: 6
  • US: 5
  • China, Spain, Taiwan, Korea, Russia, India, Hungary, Mexico: 1

Why did we organize this contest?

For the following reasons:

Follow us on Twitter and Facebook to be informed of our next challenge!

And congrats again to all participants!