Microsoft


Microsoft unveils Project Brainwave for real-time AI - Microsoft Research

#artificialintelligence

First, we have defined highly customized, narrow-precision data types that increase performance without real losses in model accuracy. Third, Project Brainwave incorporates a software stack designed to support the wide range of popular deep learning frameworks. Companies and researchers building DNN accelerators often show performance demos using convolutional neural networks (CNNs). Running on Stratix 10, Project Brainwave thus achieves unprecedented levels of demonstrated real-time AI performance on extremely challenging models.


Machines Taught by Photos Learn a Sexist View of Women

#artificialintelligence

If a photo set generally associated women with cooking, software trained by studying those photos and their labels created an even stronger association. Last summer, researchers from Boston University and Microsoft showed that software trained on text collected from Google News reproduced gender biases well documented in humans. Both datasets contain many more images of men than women, and the objects and activities depicted with different genders show what the researchers call "significant" gender bias. The researchers devised a way to neutralize this amplification phenomenon--effectively forcing learning software to reflect its training data.


Machines Learn a Biased View of Women

#artificialintelligence

If a photo set generally associated women with cooking, software trained by studying those photos and their labels created an even stronger association. Last summer, researchers from Boston University and Microsoft showed that software trained on text collected from Google News reproduced gender biases well documented in humans. Both datasets contain many more images of men than women, and the objects and activities depicted with different genders show what the researchers call "significant" gender bias. The researchers devised a way to neutralize this amplification phenomenon--effectively forcing learning software to reflect its training data.


Microsoft's AI is getting crazily good at speech recognition

#artificialintelligence

Microsoft's speech recognition efforts have hit a significant milestone. It can now transcribe human speech with a 5.1% error rate, Microsoft technical fellow Xuedong Huang wrote in a blog post -- the same error rate as humans. "Reaching human parity with an accuracy on par with humans has been a research goal for the last 25 years," Xuedong Huang wrote. "such as achieving human levels of recognition in noisy environments with distant microphones, in recognizing accented speech, or speaking styles and languages for which only limited training data is available."


How a new wave of machine learning will impact today's enterprise 7wData

#artificialintelligence

Advances in deep learning and other Machine Learning algorithms are currently causing a tectonic shift in the technology landscape. Betting big on an AI future, cloud providers are investing resources to simplify and promote machine learning to win new cloud customers. First, advances in computing technology (GPU chips and cloud computing, in particular) are enabling engineers to solve problems in ways that weren't possible before. For example, chipmaker NVIDIA has been ramping up production of GPU processors designed specifically to accelerate machine learning, and cloud providers like Microsoft and Google have been using them in their machine learning services.


Microsoft using AI to let gliders take decisions in air

#artificialintelligence

As its rivals get busy in developing self-driving cars, Microsoft is using Artificial Intelligence (AI) to empower autonomous gliders take decisions while they are aloft and has conducted a successful flight test in the US state of Nevada. According to a report in The New York Times late on Wednesday, Ashish Kapoor, an Indian-origin Principal Researcher at Microsoft, is leading a project in which his team tested two gliders designed to navigate the skies on their own. "Guided by computer algorithms that learned from onboard sensors, predicted air patterns and planned a route forward, these gliders could seek out thermals – columns of rising hot air – and use them to stay aloft," the report added. According to Mykel Kochenderfer, Stanford University professor of aeronautics and astronautics, Microsoft's project is a step towards self-driving vehicles "that are nimble enough to handle all the unexpected behavior that human drivers, bicyclists and pedestrians bring to public roads".


AIs that learn from photos become sexist

Daily Mail

In the fourth example, the person pictured is labeled'woman' even though it is clearly a man because of sexist biases in the set that associate kitchens with women Researchers tested two of the largest collections of photos used to train image recognition AIs and discovered that sexism was rampant. However, they AIs associated men with stereotypically masculine activities like sports, hunting, and coaching, as well as objects sch as sporting equipment. 'For example, the activity cooking is over 33 percent more likely to involve females than males in a training set, and a trained model further amplifies the disparity to 68 percent at test time,' reads the paper, titled'Men Also Like Shopping,' which published as part of the 2017 Conference on Empirical Methods on Natural Language Processing. A user shared a photo depicting another scenario in which technology failed to detect darker skin, writing'reminds me of this failed beta test Princeton University conducted a word associate task with the algorithm GloVe, an unsupervised AI that uses online text to understand human language.


Microsoft hits new record for AI speech recognition

#artificialintelligence

Microsoft recently reached a new milestone in its ability to recognize conversational speech, achieving a 5.1% word error rate (WER). Using Switchboard, speech recognition systems are tasked with transcribing conversations about topics such as politics or sports, for example. Microsoft's speech recognition capabilities are based on neural networks, and other artificial intelligence (AI) technologies. More information on Microsoft's speech recognition technology can be found in this technical report.


Machines Learn a Biased View of Women

WIRED

If a photo set generally associated women with cooking, software trained by studying those photos and their labels created an even stronger association. Last summer, researchers from Boston University and Microsoft showed that software trained on text collected from Google News reproduced gender biases well documented in humans. Both datasets contain many more images of men than women, and the objects and activities depicted with different genders show what the researchers call "significant" gender bias. The researchers devised a way to neutralize this amplification phenomenon--effectively forcing learning software to reflect its training data.


Microsoft's Cortana boosts speech recognition accuracy

Daily Mail

Some problems still to be addressed by the software include achieving human levels of recognition in noisy environments with distant microphones as well as recognising accented speech or speaking styles and languages for which only limited training data is available. Previous research has shown that humans achieve higher levels of agreement on the precise words spoken as they expend more care and effort, as in the case of professional transcribers. Previous research has shown that humans achieve higher levels of agreement on the precise words spoken as they expend more care and effort, as in the case of professional transcribers. This image shows the firm's voice translation service '[This includes] achieving human levels of recognition in noisy environments with distant microphones, in recognising accented speech, or speaking styles and languages for which only limited training data is available.