Google's secretive British DeepMind division is teaching its AI to talk like a human. The groundbreaking project has already halved the quality gap between computer systems and human speech, its creator say. Called WaveNet, it is capable of creating natural-sounding synthesized speech by analyzing sound waves from the human voice - rather than focusing on the human language. Google's DeepMind claims to have an AI that produces more natural-sounding synthesized speech. Google acquired UK-based DeepMind in 2014 for 533 million, and it has since beat a professional human Go player, learned how to play the Atari game Space Invaders and has read through thousands of Daily Mail and CNN articles.

I am gonna give a talk on introduction to deep learning. I wish to try to answer some elementary / basic questions that a novice in the field might have. I also would like to point out some caveats in DL. So, what questions and topics should I focus on? I shall put your references in my talk too.

Researchers at Google's DeepMind artificial intelligence division claim to have come up with a way of producing much more natural-sounding synthesized speech, compared with the techniques that are currently in use. Existing text-to-speech (TTS) systems tend to use a system called concatenative TTS, where the audio is generated by recombining fragments of recorded speech. There's also a technique called parametric TTS that generates speech by passing information through a vocoder, but that sounds even less natural. So DeepMind has come up with a new technique called WaveNet that learns from the audio it's fed, and produces raw audio sample-by-sample. To give an idea of how detailed that is, we're talking at least 16,000 samples per second.

Google's DeepMind unit, which is working to develop super-intelligent computers, has created a system for machine-generated speech that it says outperforms existing technology by 50 percent. U.K.-based DeepMind, which Google acquired for about 400 million pounds ( 533 million) in 2014, developed an artificial intelligence called WaveNet that can mimic human speech by learning how to form the individual sound waves a human voice creates, it said in a blog post Friday. In blind tests for U.S. English and Mandarin Chinese, human listeners found WaveNet-generated speech sounded more natural than that created with any of Google's existing text-to-speech programs, which are based on different technologies. WaveNet still underperformed recordings of actual human speech. Many computer-generated speech programs work by using a large data set of short recordings of a single human speaker and then combining these speech fragments to form new words.

Anyone that might be concerned about computers taking over look away now, because they are a step closer to sounding just like humans. Researchers in the UK at Google's DeepMind unit have been working on making computer-generated speech sound as "natural" as humans. The technology, called WaveNet, which is focused on the area of speech synthesis, or text-to-speech, was found to sound more natural than any of Google's products. However, this was only achieved after the WaveNet artificial neural network was trained to produce English and Chinese speech which required copious amounts of computing power, so the technology probably won't be hitting the mainstream any time soon. Using a convolutional neural network, which is used for artificial intelligence in deep learning, it is trained on data and then the systems make inferences about new data, in addition to being used to generate new data.

Google DeepMind claims to have significantly improved computer-generated speech with its AI technology, paving the way forward for sophisticated talking machines like those seen in sci-fi films like "Her" and "Ex-Machina." The London-based research lab, acquired by Google in 2014 for a reported 400 million, announced on Thursday that it has developed a talking computer programme called "WaveNet" that halves the quality gap that currently exists between human speech and computer speech. Although WaveNet sounds more like a human voice than existing artificial voice generators -- known as "text-to-speech" (TTS) systems -- it requires too much computing power to make it practical, meaning Google won't be integrating it into its products any time soon, according to The Financial Times. Aäron van den Oord, a research scientist, at DeepMind said: "Mimicking realistic speech has always been a major challenge, with state-of-the-art systems, composed of a complicated and long pipeline of modules, still lagging behind real human speech. Our research shows that not only can neural networks learn how to generate speech, but they can already close the gap with human performance by over 50%.

Google's DeepMind unit, which is working to develop super-intelligent computers, has created a system for machine-generated speech that it says outperforms existing technology by 50 percent. U.K.-based DeepMind, which Google acquired for about 400 million pounds ( 533 million) in 2014, developed an artificial intelligence called WaveNet that can mimic human speech by learning how to form the individual sound waves a human voice creates, it said in a blog post Friday. In blind tests for U.S. English and Mandarin Chinese, human listeners found WaveNet-generated speech sounded more natural than that created with any of Google's existing text-to-speech programs, which are based on different technologies. WaveNet still underperformed recordings of actual human speech. Many computer-generated speech programs work by using a large data set of short recordings of a single human speaker and then combining these speech fragments to form new words.

Songbirds provide a model system that neuroscientists use to understand how the brain learns and controls speech and similar skills. Much like infants learning to speak from their parents, songbirds learn their song from a tutor and practice it millions of times before reaching maturity. Also like humans, songbirds have evolved special brain regions for learning and producing their vocalizations. These newly-evolved brain regions in songbirds, known as the song system, are found within broader brain areas shared by birds and humans across evolution. So by studying how the song system works, we can learn about our own brains.

Last month I was reading about Autoencoders for Collaborative filtering. Using Autoencoders for Collaborative filtering is a fairly recent idea and it proven to be very effective beating all state-of-art SVD based methods. Actually the idea is pretty simple, instead of using linear function for Matrix Factorization (dot product of the two latent factor matrices, use some complex non-linear function of the two matrices so as to capture more complicated dependencies. Autoencoders are a way to do so. Matrix Factorization based CF minimizes the following objective: Note that in the objective function only known/observed user-rating pairs (set O) are considered.