Problems such as predicting a new shading field (Y) for an image (X) are ambiguous: many very distinct solutions are good. Representing this ambiguity requires building a conditional model P (Y X) of the prediction, conditioned on the image. Such a model is difficult to train, because we do not usually have training data containing many different shadings for the same image. As a result, we need different training examples to share data to produce good models. This presents a danger we call "code space collapse" -- the training procedure produces a model that has a very good loss score, but which represents the conditional distribution poorly. We demonstrate an improved method for building conditional models by exploiting a metric constraint on training data that prevents code space collapse. We demonstrate our model on two example tasks using real data: image saturation adjustment, image relighting. We describe quantitative metrics to evaluate ambiguous generation results. Our results quantitatively and qualitatively outperform different strong baselines.

Natural Language Inference is an important task for Natural Language Understanding. It is concerned with classifying the logical relation between two sentences. In this paper, we propose several text generative neural networks for generating text hypothesis, which allows construction of new Natural Language Inference datasets. To evaluate the models, we propose a new metric -- the accuracy of the classifier trained on the generated dataset. The accuracy obtained by our best generative model is only 2.7% lower than the accuracy of the classifier trained on the original, human crafted dataset. Furthermore, the best generated dataset combined with the original dataset achieves the highest accuracy. The best model learns a mapping embedding for each training example. By comparing various metrics we show that datasets that obtain higher ROUGE or METEOR scores do not necessarily yield higher classification accuracies. We also provide analysis of what are the characteristics of a good dataset including the distinguishability of the generated datasets from the original one.

Our finding continues the modern trend of achieving strong results with decades-old ideas. For example, in 2012, the "AlexNet" paper showed how to design, scale and train convolutional neural networks (CNNs) to achieve extremely strong results on image recognition tasks, at a time when most researchers thought that CNNs were not a promising approach to computer vision. Similarly, in 2013, the Deep Q-Learning paper showed how to combine Q-Learning with CNNs to successfully solve Atari games, reinvigorating RL as a research field with exciting experimental (rather than theoretical) results. Likewise, our work demonstrates that ES achieves strong performance on RL benchmarks, dispelling the common belief that ES methods are impossible to apply to high dimensional problems. ES is easy to implement and scale.

How is a network trained? When given input data with a labeled answer (meaning it's already been classified), the deviation between the network's predictions and the actual answer produces an error signal. The error signal and non-linearity of each neuron's decision function tells us whether to increase or decrease each weight. The error signal gets propagated backwards all the way to the lowest layer. Over many training examples, the network weights are repeatedly tuned until finally reaching some satisfactory benchmark, such as accuracy level.

There have been many publications in various fields from Google over the years but the AI publications trend shows how Google is trying to transform it to an AI company in future. Google publications include several fields like climate change, neurosciences, computer vision, computer games,etc. Google's growing investment in artificial intelligence, particularly deep learning, a technique whose ability to make sense of images and other data is enhancing services like search, translation and speech detection. According to Technology review, it published 218 journal or conference papers on machine learning in 2016, nearly twice as many as it did two years ago. Google also hiring many Machine learning and AI researches and almost tripled its researchers headcount over few years.

This month OpenAI published a paper "Evolution Strategies as a Scalable Alternative to Reinforcement Learning" by Tim Salimans, Jonathan Ho, Xi Chen, Ilya Sutskever which shows Evolution Strategies (ES) can be a strong alternative to Reinforcement Learning (RL) and have a number of advantages like ease of implementation, invariance to the length of the episode and settings with sparse rewards, better exploration behaviour than policy gradient methods, ease to scale in a distributed setting. Running on a computing cluster of 80 machines and 1,440 CPU cores, authors' implementation was able to train a 3D MuJoCo humanoid walker in only 10 minutes (A3C on 32 cores takes about 10 hours). Using 720 cores they can also obtain comparable performance to A3C on Atari while cutting down the training time from 1 day to 1 hour. The communication overhead of implementing ES in a distributed setting is lower than for reinforcement learning methods such as policy gradients and Q-learning. By not requiring backpropagation, black box optimizers (the ones make no assumptions about the structure of the function being optimized) reduce the amount of computation per episode by about two thirds, and memory by potentially much more.

At SVDS, our R&D team has been investigating different deep learning technologies, from recognizing images of trains to speech recognition. We needed to build a pipeline for ingesting data, creating a model, and evaluating the model performance. However, when we researched what technologies were available, we could not find a concise summary document to reference for starting a new deep learning project. One way to give back to the open source community that provides us with tools is to help others evaluate and choose those tools in a way that takes advantage of our experience. We offer the chart below, along with explanations of the various criteria upon which we based our decisions.

These are three terms that are heard all the time now, but often people still get confused about what each one really entails. Below is a quick rundown of each that will hopefully things out a little and give you a real insight as to what these interchangeable terms mean.

The deal between Google and the U.K. National Health Service was profiled by Digital Journal last November. The agreement centered on plans to develop a platform capable of sharing patient data with the aim of improving patient outcomes. This was by providing information about medical conditions with the aid of artificial intelligence. A secondary aim was to reduce the amount of paperwork by digitizing patient records. One aspect of the project involve sharing some million patient records, provided by London's Royal Free Hospital, with DeepMind.

In 2011, scientists made one of the most important discoveries in the history of AI development. They found that graphics processing units (GPUs) are far better at simulating biological learning than central processing units (CPUs). In retrospect, it seems obvious. Human brains are much more like GPUs than CPUs. Both brains and GPUs rely on parallel processing that simulates and predicts real world physics.