AI Weekly 21 Oct 2017

3rd AI Weekly is here! Google published couple interesting posts, also David Silver & Julian Schrittwieser from DeepMind answered lot of tricky questions during reddit AMA. This week we extended weekly with links to training sets, hopefully you will like it. Enjoy!

1. Pixel Visual Core: image processing and machine learning on Pixel 2 — The camera on the new Pixel 2 is packed full of great hardware, software and machine learning (ML), so all you need to do is point and shoot to take amazing photos and videos. One of the technologies that helps you take great photos is HDR+, which makes it possible to get excellent photos of scenes with a large range of brightness levels, from dimly lit landscapes to a very sunny sky.

2. Portrait mode on the Pixel 2 and Pixel 2 XL smartphones — Portrait mode, a major feature of the new Pixel 2 and Pixel 2 XL smartphones, allows anyone to take professional-looking shallow depth-of-field images. This feature works with both the rear-facing and front-facing cameras, even though neither is dual-camera (normally required to obtain this effect). Today we discuss the machine learning and computational photography techniques behind this feature.

3. A brief guide to mobile AI chips — Do I need one? What is it? Seriously, what’s going on?

4. Intel unveils new family of AI chips to take on Nvidia’s GPUs — Details are thin, but Intel says its new chips will boost deep learning training times.

5. Andrew Ng Has a Chatbot That Can Help with Depression — Woebot combines cognitive behavior therapy with advances in natural language to create a virtual counselor.

6. Reddit AMA with David Silver & Julian Schrittwieser from DeepMind, AlphaGo team.

7. THE NEURAL NET TANK URBAN LEGEND — AI folklore tells a story about a neural network trained to detect tanks which instead learned to detect time of day; investigating, this probably never happened.

1. Generalization in Deep Learning — This paper explains why deep learning can generalize well, despite large capacity and possible algorithmic instability, nonrobustness, and sharp minima, effectively addressing an open problem in the literature. Based on our theoretical insight, this paper also proposes a family of new regularization methods. Its simplest member was empirically shown to improve base models and achieve state-of-the-art performance on MNIST and CIFAR-10 benchmarks. Moreover, this paper presents both data-dependent and data-independent generalization guarantees with improved convergence rates. Our results suggest several new open areas of research.

2. AlphaGo Zero: Learning from scratch — The paper introduces AlphaGo Zero, the latest evolution of AlphaGo, the first computer program to defeat a world champion at the ancient Chinese game of Go. Zero is even more powerful and is arguably the strongest Go player in history.

3. Generalizing from Simulation — Physical robots, with controllers trained in simulation, which react to unplanned changes during simple tasks.

1. Street View Image, Pose, and 3D Cities Dataset — The repo of Street View Image, Pose, and 3D Cities Dataset. Used in “Generic 3D Representation via Pose Estimation and Matching”, ECCV16

2. Announcing AVA: A Finely Labeled Video Dataset for Human Action Understanding — AVA consists of URLs for publicly available videos from YouTube, annotated with a set of 80 atomic actions (e.g. “walk”, “kick (an object)”, “shake hands”) that are spatial-temporally localized, resulting in 57.6k video segments, 96k labeled humans performing actions, and a total of 210k action labels.

1. Meet Horovod: Uber’s Open Source Distributed Deep Learning Framework for TensorFlow

If you want to be always on time with AI weekly, feel free to follow it’s page on fb