f(AI)nder View
August 2018
f(AI)nder View - a monthly digest in the field of
Artificial Intelligence, Machine Learning, Deep Learning
exploring the various levels and faces of AI: from basic to sophisticated ones,
from algorithms and technologies to business applications

AI ecosystem
AI industry analytics, trends, foresights, insights, challenges, regulations, etc.
What's the real value of AI for your business and how can you capitalise?
$15,7 trillion - Potential contribution to the global economy by 2030 from AI.
Up to 26% boost in GDP for local economies from AI by 2030.
Read more
USA is the clear world market leader for AI, China is number two. Silicon Valley area is the world's largest AI hub, followed by London, Tel Aviv, New York, and then Beijing. Boston, Tokyo, Shanghai, Los Angeles, and Paris are still in the Top Ten 10 for global AI cities.
Read more
Technologies will transform the nature of work and the workplace itself. As a result, some occupations will decline, others will grow, and many more will change.
How AI and automation will affect work? Executive briefing from the McKinsey
Global Institute.
Read more
The function which gives AI value – is the ability to make predictions. What's going to happen is that these prediction machines are going to make predictions better and faster and cheaper. So what does this actually all mean, for us as humans?
Read more
Sundar Pichai, CEO Google: "We recognize that such powerful technology raises equally powerful questions about its use. We will assess AI applications in view of the following objectives."
Read more
Facebook's chief AI scientist, Yann LeCun, says that letting AI experts split their time between academia and industry is helping drive innovation.
Read more
Applications
What a RL agent does is just trial-and-error: it learns how good or bad its actions are based on the rewards it receives from the environment. And this is exactly how human learns to make a decision.
Read more
The technique is based on a form of machine intelligence known as a generative adversarial network. This consists of two neural networks called a generator and a discriminator.
Read more
The passport-free facial recognition system confirms a traveller's identity by matching his or her face against stored data.
And soon, you'll be able to enter a country without carrying a physical passport -- or even saying a word to another human being.
Read more
Personalization is especially important in autism therapy. MIT research team realized that deep learning would help the therapy robots perceive the children's behavior more naturally. The researchers built a personalized framework that could learn from data collected on each individual child.
Read more
Machine learning models have been successfully applied to speed up the materials discovery process. The authors propose a novel approach to both simplify the current paradigm and accelerate development. The new "closing the loop" paradigm can collect feedback immediately and optimize the exploration process.
Read more
CosmoFlow is the first large-scale science application of the TensorFlow framework at supercomputer scale with fully-synchronous training. It enables to process large 3D dark matter distribution and predict the cosmological parameters with unprecedented accuracy process.
Read more
Introductions
Neural style transfer is an optimization technique used to take three images, a content image, a style reference image (such as an artwork by a famous painter), and the input image you want to style — and blend them together such that the input image is transformed to look like the content image, but "painted" in the style of the style image.
Read more
Glow models can generate realistic-looking high-resolution images, and can do so efficiently. Glow is a type of reversible generative model, also called flow-based generative model. Accurate generative models have broad applications, including speech synthesis, text analysis and synthesis, semi-supervised learning and model-based control.
Read more
Mastering the strategy, tactical understanding, and team play involved in multiplayer video games represents a critical challenge for AI research.
Through new developments in reinforcement learning agents have achieved human-level performance in a complex multi-agent environment and one of the canonical 3D first-person multiplayer games.
Read more
Martin Gorner: "To help more developers embrace deep-learning techniques, without the need to earn a Ph.D., I have attempted to flatten the learning curve by building a short crash-course (3 hours total).
The course is focused on a few basic network architectures, including dense, convolutional and recurrent networks, and training techniques".
Read more
Generative adversarial networks (GANs) are a class of neural networks that are used in unsupervised machine learning.
GANs can be applied in many fields from generating images to predicting drugs. Unsupervised learning is a next frontier in artificial intelligence and we are moving towards it.
Read more
If there is one tool which every data scientist should use or must be comfortable with, it is Jupyter Notebooks (previously known as iPython notebooks as well).
Jupyter Notebooks are powerful, versatile, shareable and provide the ability to perform data visualization in the same environment.
Read more
Toolbox
Dactyl learns from scratch using the same general-purpose reinforcement learning algorithm and code as OpenAI Five. Reorienting an object in the hand requires the following problems to be solved: Working in the real world, High-dimensional control, Noisy and partial observations, Manipulating more than one object.
Dactyl learns to solve the object reorientation task entirely in simulation without any human input.
Read more
The ultimate goal of AutoML is to allow domain experts with limited data science or machine learning background easily accessible to deep learning models.
Auto-Keras provides functions to automatically search for architecture and hyperparameters of deep learning models.
It is developed by DATA Lab at Texas A&M University and community contributors.
Read more
Automated neural architecture search approach for designing mobile models using reinforcement learning. The overall flow of the approach consists mainly of three components: a RNN-based controller for learning and sampling model architectures, a trainer that builds and trains models to obtain the accuracy, and an inference engine for measuring the model speed on real mobile phones using TensorFlow Lite.
Read more
There are two basic concepts in reinforcement learning: the environment (namely, the outside world) and the agent (namely, the algorithm you are writing).
Gym is an open-source library, which gives you access to a standardized set of environments. It supports teaching agents everything from walking to playing games like Pong or Pinball.
Read more
It is proposed an architecture that represents numerical quantities as linear activations which are manipulated using primitive arithmetic operators, controlled by learned gates.
Experiments show that NALU-enhanced neural networks can learn to track time, perform arithmetic over images of numbers, translate numerical language into real-valued scalars, execute computer code, and count objects in images.
Read more
Visual and audio events tend to occur together: a musician plucking guitar strings and the resulting melody; a wine glass shattering and the accompanying crash; the roar of a motorcycle as it accelerates. What can be learnt by looking at and listening to a large number of unlabelled videos?
Read more
AI hardware
Traditional chip makers are putting an increasing focus on AI chip development, venture capital is pumping significant investments into the market, and AI chip startups are emerging. Machine learning technology places great demands on computing power for training algorithms and running applications, which traditional computing hardware cannot provide.
Read more
GPUs are already not commoditized relative to CPUs, and what we're seeing with the huge surge of investment in AI chips is that GPUs will ultimately be replaced by something even more specialized.
The most likely candidate is ASICs (application-specific integrated circuit), which can be highly optimized to perform a specific task.
Read more
IC Vendors (Intel, Qualcomm, NVIDIA, AMD, IBM, etc.), Tech Giants & HPC Vendors (Google, Amazon, Facebook, Apple, Microsoft, Baidu, Tesla, etc.), traditional IP Vendors (ARM, Synopsys, Imagination, etc.), Startups in China and Worldwide.
Read more
The global deep learning chipset market is segmented based on chip type, industry vertical, technology, and geography.
Top impacting factors: Increase in demand for smart homes and smart cities, Increased adoption of Deep Learning Chips in the developing regions, Rise in investments in AI startups, Development of smarter robots.
Read more
Field Programmable Gate Array (FPGA) is a reconfigurable integrated circuit.
You can configure the FPGA to become any circuit you want to (as long as it fits on the FPGA). This is quite a bit different than the instruction-based hardware most programmers are used to, such as CPUs and GPUs.
Read more
The 3D-printed artificial neural network can be used in medicine, robotics and security. The process of creating the artificial neural network began with a computer-simulated design. Then, the researchers used a 3D printer to create very thin, 8 centimeter-square polymer wafers. Together, a series of pixelated layers functions as an "optical network".
Read more
AI competitions
The Conversational Intelligence Challenge 2
The aim of the competition is to establish a concrete scenario for testing chatbots that aim to engage humans, and become a standard evaluation tool in order to make such systems directly comparable.
End date = Sep 30th.
NIPS 2018 Competition.
Read more
Typical learning problems of this kind include customer relationship management, on-line advertising, recommendation, sentiment analysis, fraud detection, spam filtering, transportation monitoring, econometrics, patient monitoring, climate monitoring, manufacturing and so on.
End date = Oct 23th.
NIPS 2018 Competition.
Read more
The main purpose of the competition is to probe the frontier of the state of the art in machine learning in the interactive and embodied setting. The competition is designed to evaluate the real ability for these learning-based systems to control physical mobile robots.
Start date = Oct 1st.
NIPS 2018 Competition.
Read more
Events
Artificial Intelligence Conference
4-7 September, 2018. San Francisco, CA.
Discover how you can apply the latest AI breakthroughs and best practices to your business. Including training courses and
AI Business Summit.
Read more
18-19 September, 2018. Mountain View, CA.
The AI Hardware Summit is the first and only conference dedicated solely to the ecosystem developing hardware accelerators for neural networks and computer vision.
Read more
19-20 September, 2018. Marina Bay Sands Expo and Convention Centre, Singapore.
"APAC: The World's AI Powerhouse".
AI Summit Singapore is part of TechXLR8 Asia, a festival of technology.
Read more
Follow us
f(AI)nder View - a monthly digest in the field of
Artificial Intelligence, Machine Learning, Deep Learning
exploring the various levels and faces of AI:
from basic to sophisticated ones,
from algorithms and technologies to business applications

f(AI)nder View is prepared by f(AI)nder Lab
Twitter
Made on
Tilda