Ian Goodfellow filled the whole Great Hall at New Orleans.

TL;DR: at the bottom.

I have just returned from the International Conference on Learning Representations (ICLR) 2019 in New Orleans and what a fruitful year that was on GAN papers. In the first section, I discuss the themes featuring papers on image synthesis (BigGAN), audio (WaveGAN), feature selection (KnockoffGAN), 3D, text & tabular and many other! The second part of this article is then focused on more practical ML considerations.

FORTUNE FAVOURS THE PREPARED

Before going to ICLR, I made a list of all the talks & workshops that had something that I wanted to learn. This meant a very busy Monday — where…


Photo by W A T A R I

It started out as a regular Friday night. Friends, drinks, food & enjoying the conversations. By 7:32 PM, my whole weekend and free time for next week was gone. I had to book an international flight to deal with authorities. Had to had dozens of phone calls in the next week.

Having worked in IT in large organizations, I’m very familiar with the pain of trying to get through something that does not quite fit with the pre-existing structures — from a perspective of someone on the inside. Last weekend, I got a taste of being on the “customer” side…


For an Artificial Intelligence (AI) professional, or data scientist, the barrage of AI-marketing can evoke very different feelings than for a general audience.

For one thing, the AI industry is incredibly broad and has many different forms and functions, so industry professionals tend to focus more deeply on which branches of AI are being hyped today. This is also true because there are greatly varying degrees of progress across the different areas of AI.

Some subfields, such as computer vision, have seen only marginal progress, whereas other areas are growing by leaps and bounds every six months. One such area…


I went to Stockholm to ICML (10–15th July) and understood academese, so you don’t have to! Domain adaptation, 3D GANs, Data Inputation using GANs and much more.

Adversarial Domain adaptation based on CycleGAN — such as CyCADA — are an interesting way to use synthetic or augmented data to help training with e.g. autonomous cars.

In all seriousness, however, I do respect greatly all the amazing work that the researchers at ICML have presented. I would not be capable of anywhere near their level of work so kudos to them for pushing the field forward!

ICML overall was an awesome experience. But this piece is not going to be about my thoughts, impressions or experiences. The industry is slowly seeping into academic conferences — whether you think that’s…


Recently, I witnessed a conversation on LinkedIn between two professionals who both work in AI — albeit as non technical staff. This conversation was about Google’s AutoML, a system that can automatically optimise itself to achieve better performance at a given task. The tragedy stems from the wording that the Google research blog has used: The article calls the two networks controller and a child. Then the article went on to explain that AI designed can outperform the human made ones.

Needless to say the term child has caused a mild panic as it has been reported by several mainstream


INTRO

I’ll jump straight into what we have explained on a high-level last time. The code is also available on GitHub and on Medium. This part is identical to the Jupyter notebook, except it is lacking the code output.

Just a sanity check that my synthesising of faces works as expected.

GENERATIVE ADVERSERIAL NETWORKS & SEMI-SUPERVISED LEARNING

BY JAKUB LANGR

This code was written for me to experiment with some of the recent advancements in AI. I chose specifically semi-supervised learning and Generative Adverserial Networks (GANs) to push myself. Some of the code was done as a homework for the Creative Applications of Deep Learning Course, which was extremely helpful in helping me learn about modern AI. Some of the broad framework…


Lots of people are arguing about the theoretical nature of what AI means many years down the road, but I would like to focus on the AI commercial products in the next 2–5 years. I’d also bring all the incredible, but deeply technical advances back to the real world and business and focus on and focus on what businesses and consumers can practically expect of the future. First off I will talk about the most interesting advances in AI, second what the current applications of those advances are and wrap it up with what this means for business.

At the…

Jakub Langr

Author GANs in Action by @ManningBooks. Course designer & lecturer at @unibirmingham. Guest lecturer @uniofoxford. ML @everMudano, Startups, Improv. ex-@join_ef

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store