Deep clustering is a recently introduced deep learning architecture that uses discriminatively trained embeddings as the basis for clustering. It was recently applied to spectrogram segmentation, resulting in impressive results on speaker-independent multi-speaker separation. In this paper we extend the baseline system with an end-to-end signal approximation objective that greatly improves performance on a challenging speech separation. We first significantly improve upon the baseline system performance by incorporating better regularization, larger temporal context, and a deeper architecture, culminating in an overall improvement in signal to distortion ratio (SDR) of 10.3 dB compared to the baseline of 6.0 dB for two-speaker separation, as well as a 7.1 dB SDR improvement for three-speaker separation. We then extend the model to incorporate an enhancement layer to refine the signal estimates, and perform end-to-end training through both the clustering and enhancement stages to maximize signal fidelity. We evaluate the results using automatic speech recognition. The new signal approximation objective, combined with end-to-end training, produces unprecedented performance, reducing the word error rate (WER) from 89.1% down to 30.8%. This represents a major advancement towards solving the cocktail party problem.

What is interesting to notice, is that the network doesn't seem to have learned detailed representations of faces. In e.g. the visualization featuring the collar, the face looks more like a spooky flesh-colored blob than a face. This might be an artifact of the visualization process, but it's not entirely unlikely that the network have either not found it necessary to learn the details, or not had the capacity to learn them. There also are a surprisingly large number of units that detect dog-related features. I counted somewhere around 50, out of 512 units in the layer in total, which means a surprising 10% of the network may be dedicated solely to dogs.

The NHS is to use artificial intelligence in the fight against sight loss with a retina-scanning system that detects the early signs of eye disease. One million patient retina scans are to be fed into an AI computer to teach it how to detect age-related macular degeneration, diabetic retinopathy and other conditions. The system is being developed by Moorfields Eye Hospital, in London, and DeepMind, the British AI company owned by Google. DeepMind was criticised by privacy groups in April after it emerged that a London hospital had granted the company access to as many as 1.6 million patient records…

An eye hospital in London is entering the debate over patient privacy by sharing images of patients' retinas with a Google-owned artificial intelligence project. Moorfields Eye Hospital is anonymizing and then sharing patient information with DeepMind, a machine-learning AI company that plans to use the hospital's non-invasive retina scans to train its machines, which must scan thousands and thousands of images to "learn" how an eye should look. While patients consented to general research, privacy advocates have expressed concerns they may not have realized the extent to which their personal information – even scans of their eyes – would be handed over to an outside party such as Google. The hospital has taken an important first step by informing patients before the project begins, which is likely a lesson learned from a previous medical research project in Britain. In a previous data-sharing partnership between DeepMind and three other London hospitals, patients discovered the involvement of their data only haphazardly afterward, the BBC reported.

We are excited to announce that the third eBook in our technical blog book series, Lessons for Large-Scale Machine Learning Deployments on Apache Spark, has been released today! This eBook, the third of a series, picks up where the second book left off on the topic of advanced analytics, and jumps straight into practical tips for performance tuning and powerful integrations with other machine learning tools – including the popular deep learning framework TensorFlow and the python library scikit-learn. The second section of the book is devoted to addressing the roadblocks in developing machine learning algorithms on Apache Spark – from simple visualizations to modeling audiences with Apache Spark machine learning pipelines. As with the past eBooks, we've augmented the blogs with code examples in Databricks notebooks, which are complimentary with the eBook download. Download the eBook to get started on your next advanced analytics project today.

Across the world there is an estimated 285 million visually impaired people, and 39 million of these are blind. Conditions like age-related macular degeneration and diabetic retinopathy can be picked up is using digital screenings, which are highly complex and take a lot of time to analyse. Now Google's DeepMind Health is teaming up with a London eye hospital to investigate how machine learning could help analyse these scans efficiently and effectively. Moorfields Eue Hospital in London has announced a new medical research partnership with Google's DeepMind Health that could revolutionise the way professionals carry out eye tests and lead to earlier detection of common eye diseases Diabetes is on the rise. It's estimated that 1 in 11 of the world's adult population are affected.

Along the way, I'll introduce Deep Learning, and add context and background on why the classifier works so well. Here are links to learn more, thanks for watching, and have fun! You can follow me on Twitter at https://twitter.com/random_forests for updates on episodes, and of course - Google Developers.

Machine learning is a branch of artificial intelligence concerned with the design of data-driven programs which autonomously demonstrate intelligent behavior in a variety of domains. Machine learning systems are all around us. When you deposit a check, scan your fingerprint, or post a picture on social media, autonomous algorithms are deployed on the spot to sift through and make sense of your constant interactions with our technology. Machine learning silently underpins the fabric of our digital infrastructure, discriminating spam e-mail and banking fraud, making light-speed transactions in the global financial market, recommending music and films for customers to buy, deciding what search results are relevant to your queries, and countless more of the daily interactions with electronic media that we take for granted. Machine learning is the backbone that powers self-driving cars, content recommendation in social media, face identification in digital forensics, and countless other high-level tasks.

DeepMind, Google's London-based artificial intelligence company, has started training neural networks to recognise the signs of eye disease in medical images. A partnership with Moorfields Eye Hospital in London has given the company access to about a million anonymised retinal scans, which DeepMind will feed into its artificial intelligence software. The project will target two of the most common eye diseases – age related macular degeneration and diabetic retinopathy. More than 100 million people around the world have these conditions. The information that Moorfields is providing includes scans of the back of people's eyes, as well as more detailed scans known as optical coherence tomography (OCT). The idea is that the images will let DeepMind's neural networks learn to recognise subtle signs of degenerating eye conditions that even trained clinicians have trouble spotting.

Deep learning has become of the most relevant trends in modern software technology. From a conceptual standpoint, deep learning is a discipline of machine learning that focuses on modeling data using connected graphs with multiple processing layers. In the last few years, deep learning has become a pivotal technology to power uses cases such as image recognition, natural language processing or even powering some of the capabilities of self-driving vehicles. The popularity of deep learning has expanded beyond just software and now the industry is starting to talk about the first generation of hardware with deep learning capabilities: a deep learning chip. A few months ago, at its I/O Conference, Google announced the design of an application-specific integrated circuit (ASIC) focused on deep learning capabilities and neural nets.