Google Releases Pixel 2 Portrait Mode Deep Learning Model – Dr. Rami Shaheen

Many of Google’s machine learning efforts are open-sourced so that developers can take advantage of the latest advancements. The latest release is for semantic image segmentation or the technology behind the Pixel 2’s single lens portrait mode.This deep learning model assigns semantic labels to every pixel in an image. In turn, categorization allows classifications like road, sky, person, or dog, and which part of a picture is the background and what is the foreground.Applied to photography, the latter is leveraged on the Pixel 2’s Portrait Mode for shallow depth-of-field effects with only one physical lens. This use requires optimization especially in “pinpointing the outline of objects,” or being able to distinguish where a person ends and the background begins.Assigning these semantic labels requires pinpointing the outline of objects, and thus imposes much stricter localization accuracy requirements than other visual entity recognition tasks such as image-level classification or bounding box-level detection.

2021-01-20T08:41:00+00:00January 20th, 2021|

YouTuber challenges scientist to create an AI version of him for $100

We already know deepfakes are all over the place these days, and the technology associated with them is advancing rapidly. But how easy is it to create a digital replica of somebody? And could it be done on a budget? That's the question YouTuber Tom Scott set out to find the answer to in his latest video, for which he challenged AI in neuroscience researcher Jordan Harrod to create a fake version of him for $100. "This isn't a face replacement or a body double," Scott explains at the start of the video. "That's mainstream now. Hollywood's doing it, anyone can do it. This is trying to create a cheaper, complete replacement for me — an automatic digital puppet on a budget of just $100." As you can probably guess, Harrod manages this easily enough — creating a version of Scott that, while a little rough around the edges, still manages to help her narrate the rest of the video while doing a passable impression of Scott's voice and facial expressions. But rather than being an instructional video (which, [...]

2021-01-20T04:40:27+00:00January 20th, 2021|

Facebook and Instagram’s AI-generated image captions now offer far more details – Dr. Rami Shaheen

Every picture posted to Facebook and Instagram gets a caption generated by an image analysis AI, and that AI just got a lot smarter. The improved system should be a treat for visually impaired users, and may help you find your photos faster in the future. Alt text is a field in an image’s metadata that describes its contents: “A person standing in a field with a horse,” or “a dog on a boat.” This lets the image be understood by people who can’t see it. These descriptions are often added manually by a photographer or publication, but people uploading photos to social media generally don’t bother, if they even have the option. So the relatively recent ability to automatically generate one — the technology has only just gotten good enough in the last couple years — has been extremely helpful in making social media more accessible in general. Facebook created its Automatic Alt Text system in 2016, which is eons ago in the field of machine learning. The team has since cooked up many improvements to it, making [...]

2021-01-20T00:34:26+00:00January 20th, 2021|

System brings deep learning to “internet of things” devices | MIT News

Deep learning is everywhere. This branch of artificial intelligence curates your social media and serves your Google search results. Soon, deep learning could also check your vitals or set your thermostat. MIT researchers have developed a system that could bring deep learning neural networks to new — and much smaller — places, like the tiny computer chips in wearable medical devices, household appliances, and the 250 billion other objects that constitute the “internet of things” (IoT). The system, called MCUNet, designs compact neural networks that deliver unprecedented speed and accuracy for deep learning on IoT devices, despite limited memory and processing power. The technology could facilitate the expansion of the IoT universe while saving energy and improving data security. The research will be presented at next month’s Conference on Neural Information Processing Systems. The lead author is Ji Lin, a PhD student in Song Han’s lab in MIT’s Department of Electrical Engineering and Computer Science. Co-authors include Han and Yujun Lin of MIT, Wei-Ming Chen of MIT and National University Taiwan, and John Cohn and Chuang Gan of the [...]

2021-01-19T20:32:19+00:00January 19th, 2021|

US gov sets up the National Artificial Intelligence Initiative Office at the last minute before Trump’s presidency ends – Dr. Rami Shaheen

In brief Just days before President Trump is due to leave office, the White House created its National Artificial Intelligence Initiative Office, which will sit under the Office of Science and Technology Policy. The org will be in charge of helping industry and academia work together to ensure the development of AI falls in line with US strategy. It will also help pass new policies and establish R&D funding initiatives. The move is one of Trump’s last-ditch efforts to establish something potentially long lasting in the field during his time in government. “Through the Trump Administration’s historic efforts and the unparalleled enthusiasm and activity of the private sector and academia, the United States remains the world leader in artificial intelligence,” according to a statement. “The White House’s new National AI Initiative Office will be integral to the Federal Government’s efforts to maintain this leadership position for many years to come.” Facial-recognition tool used to screen job candidates has been scrapped HireVue, a software biz that developed one of the most widely used job screening algorithms, has dropped its controversial [...]

2021-01-19T16:30:15+00:00January 19th, 2021|

NHS will use AI to improve COVID-19 treatment and shorten hospital stays

The UK’s NHS (National Health Service) will use AI to help improve the outcome for COVID-19 patients and reduce their time spent in hospital. While the UK’s vaccine rollout is among the fastest in the world – certainly in comparison to its European peers – a series of late lockdowns and initial lack of PPE equipment has put the health service under immense pressure and left the country with one of the highest COVID-19 deaths rates per capita. The latest lockdown was imposed after ministers were warned the NHS would collapse within a few weeks, leaving Britons unable to access care for any reason. Paramedics have already reported patients dying in ambulances queued outside of hospitals and research from King’s College London suggests around half of ICU staff are suffering from PTSD and problem drinking due to the scenes they’ve witnessed and pressure they’re under. NHSX, the digital innovation arm of the health service, is now providing access to the National COVID-19 Chest Imaging Database (NCCID) to hospitals and universities across the country to help speed up the diagnosis [...]

2021-01-19T08:16:23+00:00January 19th, 2021|

AI Research at Amazon: Brand Voice, Entanglement Frontier, Humor Recognition  

AI research at Amazon in 2020 produced Brand Voice, explored the ‘entanglement frontier’ of quantum computing, and attempted to model attempts at humor from customers. (Photo by Christian Wiediger on Unsplash)  By John P. Desmond, - Dr. Rami Shaheen Editor    The most popular articles, blog posts, and downloaded publications of 2020 published on the Amazon Science website can serve as an update on AI work at Amazon.  The most popular article of 2020 was an account of advances in text-to-speech technologies. Amazon’s Alexa voice service has been on the market for more than five years and is now available on hundreds of millions of devices from Amazon and other device manufacturers.    Advances in text-to-speech (TTS) technologies—the ability of computers to convert sequences of words into natural sounding, intelligible audio responses—have made it possible for computers to sound more human-like.  Amazon scientists and engineers are working on predicting how the sentiment of an utterance might sound to an average listener, for example, and respond with human-like intonations.   An advance in the field occurred in 2016, when researchers at DeepMind, the London-based AI firm owned by Google, introduced WaveNet, [...]

2021-01-19T04:13:12+00:00January 19th, 2021|

Amazon SageMaker AI Service – Dr. Rami Shaheen

Amazon revealed its SageMaker AI service today, allowing its customers to train machine learning models at massive scale while keeping costs down. Amazon uses novel techniques to keep the required compute power locked down while providing comparable performance.When SageMaker takes in data to train a model, it uses a streaming algorithm that only makes one pass over the data that it gets fed. While other algorithms can see exponential increases in the amount of time and processing power needed, Amazon’s algorithms don’t. As data is streamed into the system, the algorithm adjusts its state — a persistent representation of the statistical patterns present in the information fed into SageMaker for training a particular system.That state isn’t a trained machine learning model, though: It’s an abstraction of the data fed to SageMaker that can then be used to train a model. That provides a number of useful advantages, like making it easier for Amazon to distribute training of a model. SageMaker can compare the states of the same algorithms working on different data across multiple machines over the course of [...]

2021-01-19T00:12:06+00:00January 19th, 2021|

Descript raises $30M to build the next generation of video and audio editing tools – Dr. Rami Shaheen

The popularity of podcasting and online video shows no signs of slowing down, and so we continue to see a wave of creators publishing a profusion of audio and video content to fill out the airwaves. Today, a company building a platform to make that work easier and more interesting to execute is announcing a round of growth funding to double down on the opportunity. Descript, which builds tools that let creators edit audio and video files by using, for example, natural language processing to link the content to the editing of text files, has picked up $30 million in a Series B round of funding. Andrew Mason, the CEO and founder of the company, said in an interview that the plan will be to use the money to continue building out tools not just for mass-market and individual professional and amateur creators, but also, increasingly, organizations that might be using the tools for their own in-house video and audio needs, a use case that has definitely grown during the last year of global remote working. “We see ourselves… [...]

2021-01-18T16:05:06+00:00January 18th, 2021|