Usually, every month we write an article about the best and most promising AI research papers from that month. You can check them out here:
In addition to that, we list fifteen AI articles we have found amazing that month. This collection of articles should give you an overview of what happened that month in the AI industry both from technical, business and from an ethical perspective. You can check out the March edition here. Have fun!
We don’t do sales, but given the circumstances and the severity of the situation, we decided to change that. Don’t be fooled, this sale isn’t meant for profit and it’s most definitely not planned. This sale is here to help people who want to become better, learn new skills and be more productive than ever before. Our book offers are on a 50% sale.
A state-of-the-art open source chatbot
Building open-domain chatbot applications is a special field of machine learning, with a special set of challenges. This month, Facebook open-sourced its awesome chatbot Blender. In this article, you can find their “cookbook” for building great chatbot applications, as well as all necessary information about Blender itself. They proposed several approaches that are relying not only on scaling neural models, but on other “ingredients” as well.
The AI Economist: Improving Equality and Productivity with AI-Driven Tax Policies
High-income inequality is becoming a big problem in our society. It impacts economic growth as many studies find. One way to deal with it is tax. Setting up an optimal tax policy can reduce this inequality, however, it is extremely hard to find one. Find out how you can use Reinforcement Learning for this hard and specific task.
Google’s medical AI was super accurate in a lab. Real life was a different story.
This article shows how far are we from generalized AI solutions in healthcare. Lab and real environment are not the same, and using accuracy as the only metric for evaluation can give us a false sense of accomplishment. This becomes painfully clear when we work with hectic clinical environments.
Treating COVID-19: How Researchers Are Using AI to Scale Care, Find Cures, and Crowdsource Solutions
The Stanford Institute for Human-Centered Artificial Intelligence (HAI) held a virtual COVID-19 and AI Conference on April the 1st. In this conference, experts are discussing how to prevent, detect, understand and treat using AI technology. Digest of all proposals and detailed analysis of each approach can be found in this article.
The OpenAI Microscope is a collection of visualizations of every significant layer and neuron of eight important vision models
This is less of an article and more one very cool website. Guys from OpenAI created a website called Microscope and there you can find a collection of visualizations of every layer and neuron of several famous computer vision neural networks. You can go deep into AlexNet, Inception and ResNet.
A harsh critique of AI industry. Highly recommended.
Image Segmentation: Tips and Tricks from 39 Kaggle Competitions
Based on 39 Kaggle competition, this article contains all tricks you would possibly need for image segmentation. From data augmentation, over modeling the architecture, all the way to the hardware requirements.
OpenCV Age Detection with Deep Learning
I am a big fan of the PyImageSearch blog. These guys always have cool tutorials about deep learning in computer vision. In this article, they present one method for age detection using OpenCV.
AI researchers propose ‘bias bounties’ to put ethics principles into practice
As always, AI covers not only technology and research but ethics as well. It is good to keep in mind these things when you develop models because we transfer our biases to the model. Researchers from top research labs and companies in the U.S. and Europe released a toolbox for turning AI ethics principles into practice.
NVIDIA and King’s College London Announce MONAI Open Source AI Framework for Healthcare Research
MONAI is an open-source AI framework for healthcare from NVIDIA and Kin’s Collage. It is based on Ignite and PyTorch and it can process unique formats, resolutions, and specialized meta-information of medical images.
Announcing TorchServe, An Open Source Model Server for PyTorch
It is always said that TensorFlow is more popular in the industry and PyTorch is more popular within the research crowd. Why is that? One of the theories is that TensorFlow is popular in the industry because of the easier deployment of its models. This could change with TorchServe.
Facebook Open Sources Architecture for Personalized Neural Recommendation Systems
Upcoming changes to TensorFlow.js
In this article, you can see the roadmap for the TensorFlow.js. A demand for this technology grows and Google responds to these requirements.
Swift: Google’s bet on differentiable programming
You don’t like Python, but you like Data Science? Well, it seems that from now on you could use Swift. Google’s Swift for TensorFlow team built some really cool features into Swift and made it the first mainstream language with first-class language-integrated differentiable programming capabilities.
5 Applications of GANs – Video Presentations You Need To See
We looove GANs and we want you to looooove them too. So here are 5 really cool applications of them.
Building Smart Apps
Read our blog posts here.