For any media company that's producing and distributing large volumes of content, highly specific metadata is the foundation that other publisher tools will be built upon. Metadata provides a way of finding, identifying and classifying videos thus making them highly searchable and shareable. Using Deep Learning, Vilynx exposes the multi-level intelligence of video assets to enable publishers more efficient content management. Most recommendation systems rely on the relationships between the content in videos to provide appealing suggestions about which other videos you might also want to see. To solve this, some publishers have resorted to semantic analysis of content as well as basic collaborative filters to power content recommendations.

Recently the generalisation error of deep neural networks has been analysed through the PAC-Bayesian framework, for the case of fully connected layers. We adapt this approach to the convolutional setting.

Deep neural networks are able to solve tasks across a variety of domains and modalities of data. Despite many empirical successes, we lack the ability to clearly understand and interpret the learned internal mechanisms that contribute to such effective behaviors or, more critically, failure modes. In this work, we present a general method for visualizing an arbitrary neural network's inner mechanisms and their power and limitations. Our dataset-centric method produces visualizations of how a trained network attends to components of its inputs. The computed "attention masks" support improved interpretability by highlighting which input attributes are critical in determining output. We demonstrate the effectiveness of our framework on a variety of deep neural network architectures in domains from computer vision, natural language processing, and reinforcement learning. The primary contribution of our approach is an interpretable visualization of attention that provides unique insights into the network's underlying decision-making process irrespective of the data modality.

From a cognition and growth perspective, I played the game with my son and also my daughter as not only a means to have fun, but also since I figured it would be a good learning tool for them. One aspect of learning in this particular game is the effects of compression. When you compress items together, you need to think about how the physics of compression impacts other objects. In some instances, the compression would bear upon a handful of the pegs and the other pegs were not under any pressure at all. This at first was counter intuitive to my children as they initially assumed that applying pressure would cause all of the pegs to be under pressure.

Deep learning has brought about machines that can'see' the world more like humans can, and recognize language. And while deep learning was inspired by the human brain, the question remains: Does the brain actually learn this way? The answer has the potential to create more powerful artificial intelligence and unlock the mysteries of human intelligence. In a study published December 5th in eLife, CIFAR Fellow Blake Richards and his colleagues unveiled an algorithm that simulates how deep learning could work in our brains. The network shows that certain mammalian neurons have the shape and electrical properties that are well-suited for deep learning.

In previous tutorials on deep learning, I have taught how to build networks in the TensorFlow deep learning framework. There is no doubt that TensorFlow is an immensely popular deep learning framework at present, with a large community supporting it. However, there is another contending framework which I think may actually be better – it is called the Microsoft Cognitive Toolkit, or more commonly known as CNTK. Why do I believe it to be better? Two main reasons – it has a more intuitive and easy to use Python API than TensorFlow, and it is faster.

Why are deep neural networks hard to train? Appendix: Is there a simple algorithm for intelligence? If you benefit from the book, please make a small donation. I suggest $5, but you can choose the amount. Thanks to all the supporters who made the book possible, with especial thanks to Pavel Dudrenov.

Docker is the best platform to easily install Tensorflow with a GPU. This tutorial aims demonstrate this and test it on a real-time object recognition application. Docker is a tool which allows us to pull predefined images. The image we will pull contains TensorFlow and nvidia tools as well as OpenCV. The idea is to package all the necessary tools for image processing.

Guest blog post by Christopher Dole and other contributors, originally posted here. Deep Learning is one of the most revolutionary and disruptive technologies ever developed in Data Science. Essentially, this is a class of algorithms inspired by how the human brain works, and it has the ability to automate and replace most of the world's jobs. This is what enables self-driving cars to function and what allows Spotify to create very customized playlists and recommendations. This is how YouTube is able to identify faces and animals in videos and how Siri can understand and process free speech in milliseconds.

Artificial Intelligence (AI) is starting to make a dent in the multi-trillion dollar healthcare industry, from being able to identify cancers and healthcare issues with just a glance, to being able to decipher the mysteries of the human genome and figure out how much longer you have left to live, but now it has a new trick. In a new arXiv paper published by the researchers from Stanford University, who also trained their smart watches to identify when you're getting ill, the team behind the newest AI addition to healthcare explain how CheXNet, their Convolutional Neural Network achieved the feat. CheXNet was trained on a publicly available data set of more than 100,000 chest X-Rays that were annotated with information on at least fourteen different diseases. The team then had four Radiologists go through a test set of X-Rays and make diagnoses, and these were compared to the diagnoses performed by CheXNet. Not only did CheXNet beat the Radiologists at spotting Pneumonia, but once the algorithm was expanded, it proved better at identifying the other thirteen diseases as well.