Artificially Intelligent Painters: can deep learning AI create the next Mona Lisa?

Neural Style

If you have ever used Instagram or Snapchat, you are familiar with using filters that alter the brightness, saturation, contrast, and so on of your images. Neural style, a deep learning algorithm, goes beyond filters and allows you to transpose the style of one image, perhaps Van Gogh’s “Starry Night,” and apply that style onto any other image.  

Neural style, one of many models available on Somatic.io, uses a deep neural network in order to separate and recombine content and style of any two images. It is one of the first artificial neural networks (ANNs) to provide an algorithm for the creation of artistic imagery.

convolutional neural network

How Does it Work?

The model is given two input images, one that will be used for styling, the other for content. At each processing stage in the convolutional neural network’s (CNN) hierarchy, the images are broken into a set of filtered images. While the number of different filters increases along the processing hierarchy, the overall size of the filtered images is reduced, leading to a decrease in the total number of units per layer of the network.

The above figure visualizes the information at different processing stages in the CNN. The  content reconstructions from lower layers (a,b,c) are almost exact replicas of the original image. In the higher layers of the network however, the detailed pixel information is lost while the high-level structures and details remain the same (d,e). Meanwhile, the model captures the style of the other input image on top of the content CNN representations. Then, the style representation draws connections between the different features in different layers of the CNN. The model then reconstructs the style of the input image on top of the content representations within each of the CNN layers. This creates images that match the style on an increasing scale as you move through the network’s hierarchy.

convolutional neural network layers

Try It Out!

Experiment with the model for yourself. All you need to do is select an image you want to use for style and anther one for the content. Here are some creations of the latest creations the model has generated.


IT leaders’ pragmatism will be the antidote to AI scaremongering

IT leaders' pragmatism will be the antidote to #AI dystopia:  #MachineLearning #AI4socialgood

  • Every development in AI is portrayed as the forerunner to Skynet from the Terminator movies, or something from Blade Runner, Westworld or another vision of a future ruled by robot overlords.
  • Paypal founder and billionaire entrepreneur Elon Musk joined in, warning that AI represented the greatest threat to mankind.
  • Sadly, we’re probably going to have to get used to this for a few years yet, until AI becomes more mainstream, creates as many jobs as it eliminates, and starts to deliver huge benefits to businesses and society – much like new technologies have for the last 50 years.
  • We are, in many ways, starting into an age where what was once science fiction will become a reality – but just because sci-fi writers realised that dystopian visions sell more books than utopian dreams, we’ve become culturally conditioned to the idea that too much new technology is a bad thing.
  • For IT professionals, your job is to understand and explain what AI and other emerging technologies can bring to your business and your customers – and to deliver the enormous potential on offer.

Clearly we are going through the phase in the development of artificial intelligence (AI) technology where rationality and reasoned debate are replaced by science-fiction scaremongering and …
Continue reading “IT leaders’ pragmatism will be the antidote to AI scaremongering”

Large Scale Machine Learning for Payment Fraud Prevention Recorded at:

How advanced #machinelearning algorithms are applied at @PayPal for #fraud prevention. 


  • Venkatesh Ramanathan is a senior data scientist at PayPal where he is working on building state-of-the-art tools for payment fraud detection.
  • Venkatesh has worked on various problems in the areas of Anti-Spam, Phishing Detection, and Face Recognition.
  • Data Science is an emerging field that allows businesses to effectively mine historical data and better understand consumer behavior.
  • This type of scientific data management approach is critical for any business to successfully launch its products and better serve its existing markets.

Venkatesh Ramanathan presents how advanced machine learning algorithms such as Deep Learning and Gradient Boosting are applied at PayPal for fraud prevention.
Continue reading “Large Scale Machine Learning for Payment Fraud Prevention Recorded at:”

AI Superstar Andrew Ng Is Democratizing Deep Learning With A New Online Course

#AI hero @AndrewYNg is #democratizing #deeplearning w/ new online course
in @FastCompany

  • That’s the vision of Andrew Ng, a founder of the Google Brain deep learning project, and former head of AI at Baidu–a position he left in March–who is today announcing a set of five interconnected online courses on the subject.
  • “Today, if you want to learn deep learning, there are lots of people searching online, reading [dozens of] research papers, reading blog posts, and watching YouTube videos,” Ng tells Fast Company.
  • As Ng sees it, getting to an AI-powered economy is going to take the work of much more than any one, or even several companies.
  • “I hope we can build an AI-powered future that provides everyone affordable healthcare, accessible education, inexpensive and convenient transportation, and a chance for meaningful work for every man and woman,” Ng says in his announcement, which is the first from his newly created company, deeplearning.ai.
  • Ng is aware that many people are still confused by AI, often getting bogged down in the different subspecialties, and lingo that can easily be misused.

The founder of Google Brain and former head of Baidu’s AI efforts wants to train a giant new workforce to help make “AI the new electricity.”

Continue reading “AI Superstar Andrew Ng Is Democratizing Deep Learning With A New Online Course”

I blame the parents – AI needs to be raised right

I blame the parents – #AI needs to be raised right

  • Nexus CX, a pioneer in AI, is getting properly up close and personal.
  • In SU’s pilot stages, men who thought they were talking to a bot responded more openly than those who were told they were speaking with a human at the other end.
  • Nexus CX are working with Amazon’s Alexa, recording Trainor’s friend, documenting his memories and thoughts, helping to test a virtual counterpart and robot avatar that will speak based on collected patterns of speech.
  • It’ll console people in a way that humans can’t.
  • And when you think of the great technological advancements of the past decade, one creation stands head and shoulders above the rest: the iPhone.

Apparently, artificial intelligence is going to take over our lives, our jobs, our minds even and not necessarily in a good way. It’s inevitable.
Continue reading “I blame the parents – AI needs to be raised right”

Learn how AI is changing the application development landscape

Learn how #AI is changing the #application #development landscape. Get the report:

  • Is AI on your software development roadmap?
  • With technologies like advanced machine learning, deep learning, natural language processing, and business rules, AI is poised to disrupt both how developers build applications and the nature of those applications.
  • The risks—unrealistic expectations, integration with traditional applications, and more—can’t be ignored as your organization strives for the rewards of an accelerated development cycle and a new generation of self-learning applications.
  • Uncover this shifting digital landscape and how your business can take advantage of it in the Forrester Research report, “How AI Will Change Software Development And Applications.”
  • Fill out the form at right to read the free report.

Learn how AI is changing the application development landscape
Continue reading “Learn how AI is changing the application development landscape”

Accenture Business Journal for India – Vol. 3

Fuel your #IntelligentAutomation journey with a core #AI competency #ABJI2017

  • From smart connected plants and insight-driven enterprises to Blockchain-enabled services, Indian enterprises want to take digital to the next level.
  • How should they leverage the evolution of digital technology-especially in artificial intelligence-to build new business models, new products and services or enter new markets?
  • This edition of the Accenture Business Journal for India reveals the secret sauce for digital success across industries-from telecom, consumer packaged goods to manufacturing.
  • Take a deep dive and learn how to Lead in the New and avoid digital oblivion.

Accenture Business Journal for India – Vol. 3
Continue reading “Accenture Business Journal for India – Vol. 3”


Official #TensorFlow implementation of Dense Transformer Networks

  • In this work, we propose Dense Transformer Networks to apply spatial transformation to semantic prediction tasks.
  • The third and fourth rows are the segmentation results of U-Net and DTN, respectively.
  • max_epoch: how many iterations or steps to train

    test_step: how many steps to perform a mini test or validation

    save_step: how many steps to save the model

    summary_step: how many steps to save the summary

    sampledir: where to store predicted samples, please add a / at the end for convinience

    model_name: the name prefix of saved models

    test_epoch: which step to test or predict

    network_depth: how deep of the U-Net including the bottom layer

    class_num: how many classes.

  • We have conv2d for standard convolutional layer, and ipixel_cl for input pixel convolutional layer proposed in our paper.
  • We have deconv for standard deconvolutional layer, ipixel_dcl for input pixel deconvolutional layer, and pixel_dcl for pixel deconvolutional layer proposed in our paper.

Contribute to dtn development by creating an account on GitHub.
Continue reading “GitHub”