Profile cover photo
Profile photo
Wayne Radinsky
17,796 followers -
Software Design Engineer
Software Design Engineer

17,796 followers
About
Wayne's posts

Post has attachment
Fonts arranged by AI. "Once I had a bunch of font images, I used a convolutional neural network called VGG16 to generate for each font a list of numbers that represents what the network thought were the notable visual features of the image. Finally, in order to get x,y points, I ran these embeddings through T-SNE, a popular algorithm for taking large vectors and compressing them into a smaller space  --  in this case into a 2-D plane."

"At the end of the process, I had an arrangement of about 800 fonts in a 2-D space. I was pretty surprised by how well the results came out : There were very clear font clusters like a sans-serif group, and a group of cursive fonts."

Post has attachment
"Robotic construction platform creates large buildings on demand." "MIT Media Lab's Digital Construction Platform (DCP) is mobile (with a top speed of 0.5 meters per second) and self-contained. It's battery powered (with a few solar panels on it and an option for more to be attached), so it can potentially run forever, or as long as you have sun. Otherwise, the DCP mimics much of the functionality of a 3D building printer: It has a long reach, giving it a maximum printable volume of 2,786 cubic meters. The robot itself is made out of two arms, modeled loosely on a human: There's a big long arm with 4 degrees of freedom (DoF) that does all the gross motions, and one small, dexterous 6-DoF Kuka arm that takes care of fine motions like our hands and fingers would. Put it all together, and the total system cost comes to US $244,500, which is really not that bad."

Post has attachment
"Artificial intelligence is making human speech as malleable and replicable as pixels. Today, a Canadian AI startup named Lyrebird unveiled its first product: a set of algorithms the company claims can clone anyone’s voice by listening to just a single minute of sample audio."

"The results certainly aren’t indistinguishable from human speech, but they’re impressive all the same, and will no doubt improve over time. Below you can hear the synthesized voices of Donald Trump, Barack Obama, and Hillary Clinton discussing the startup."

Post has attachment
A machine learning system to detect hate speech on Twitter has been developed by Indian engineers from IIIT-Hyderabad. It uses deep learning and classifies Tweets as either 'racist' or 'sexist' as learned from its training examples.

Post has attachment
Four machine learning models ("random forest", "logistic regression", "gradient boosting", and "neural networks") have been demonstrated to be better than the algorithm currently in use at predicting a patient's future risk having of heart disease or a stroke.

"Current standard prediction models like the ACC are based on eight risk factors including age, cholesterol level and blood pressure but are too simplistic to account for other factors like medications, multiple disease conditions, and other non-traditional biomarkers. These AI algorithms have the potential to help save more lives."

The machine learning models were trained on a large data set of 378,256 patients in the UK.

Post has attachment
IBM says they have achieved an accuracy of 86% at diagnosing diabetic retinopathy using deep research. "The research found that a new method created by the IBM team achieved an accuracy score of 86 percent in classifying the severity of the disease across the five levels recognized on the international clinical DR scale (no DR; mild; moderate; severe; proliferative DR)."

"Based on more than 35,000 eye images accessed via EyePACS, the IBM technology was trained to identify lesions such as micro-aneurysms, haemorrhages and exudates to indicate damage of the retina's blood vessels and assess both the presence and severity of the disease. The novel method for classifying the severity level of DR combines deep learning techniques, convolutional neural networks (CNN), with a dictionary-based learning to incorporate DR specific pathologies."

Post has attachment
Using AI to rank chess champions and compare today's players with past players. "The ELO system, which most chess federations use today, ranks players by the results of their games. Although simple and efficient, it overlooks relevant criteria such as the quality of the moves players actually make."

"Jean-Marc Alliot's system computes the difference between the move actually played and the move that would have been selected by the best chess program to date, Stockfish. Running on the OSIRIM supercomputer, this program executes almost perfect moves. Starting with those of Wilhelm Steinitz (1836-1900), all 26,000 games played since then by chess world champions have been processed in order to create a probabilistic model for each player. For each position, the model estimates the probability of making a mistake, and the magnitude of the mistake. These models can then be used to compute the win/draw/lose probability for any given match between two players. The predictions have proven not only to be extremely close to the actual results when players have played concrete games against one another, they also fare better than those based on ELO scores. The results demonstrate that the level of chess players has been steadily increasing. The current world champion, Magnus Carlsen, tops the list, while Bobby Fischer is third."

Post has attachment
"Venezuela just did General Motors a favor by seizing its car assembly plant." "Business for GM and other foreign carmakers in the Caribbean nation was already suffering before the seizure. For years, they've struggled to import car parts and other raw materials due to government currency controls. Car buyers are also scarce; this is not surprising given that many Venezuelans can't even afford enough food." "Yet, due to the country's strict labor laws, the company has had to keep its roughly 2,700 workers on the payroll."

Post has attachment
"9 lessons from 25 years of Linux kernel development" from Greg Kroah-Hartman. Short release cycles are important, process scalability requires a distributed, hierarchical development model, tools matter, the kernel's strongly consensus-oriented model is important, the kernel's strong 'no regressions' rule is also important, corporate participation in the process is crucial, but no single company dominates kernel development, there should be no internal boundaries within the project, the kernel shows that major developments can spring from small beginnings, above all, 25 years of kernel history show that sustained, cooperative effort can bring about common resources that no group would have been able to develop on its own.

Post has attachment
How automation technology makes fraud easier for CEOs. "There is one way by which you as a CEO can not only argue that adequate controls and supervision were in place and at the same time make it easier for your employees to commit fraud. Simply perform the monitoring and control function through an automated system and restrict your role to signing off on the risk metrics that are the output of this automated system."

"Just as the algorithmic revolution is simply a continuation of the control revolution, the 'accountability gap' due to automation is simply an acceleration of trends that have been with us throughout the modern era."
Wait while more posts are being loaded