Long Short-Term Memory dramatically improves Google Voice etc – now available to a billion users

A type of “Recurrent neural networks” architecture initially developed in Juergen Schmidhuber’s research groups at the Swiss AI Lab IDSIA and TU Munich greatly improved Google Voice (by 49%) and are now available to a billion users. You can find the recent Google Research Blog on this, by Haşim Sak, Andrew Senior, Kanishka Rao, Françoise […]

Google Acquires Deep Mind

Google is acquiring an AI startup called DeepMind for more than 500 million dollars[1,2]. Deep mind has recently hired several deep learning experts and recent graduates from Geoffrey Hinton’s, Yann Lecun’s, Yoshua Bengio’s and Jurgen Schmidhuber’s groups. One of the co-founders of DeepMind, Shane Legg was a PhD student at IDSIA. According to [2] Google […]

Google’s new Deep Learning Algorithm Transcribes House Numbers

During his summer internship, Ian Goodfellow (currently a PhD student at UdeM Lisa Lab) and his collaborators from Google, Yaroslav Bulatov, Julian Ibarz, Sacha Arnoud, Vinay Shet, submitted a paper to ICLR 2014 that proposes a deep learning method which successfully transcribes the house numbers from Google Streetview images. This work took wide coverage in […]

Deep learning based Visual Search Engines

Recently Google and Baidu announced their deep learning (specifically, convolutional neural network) based visual search engines.

You can find more information related to Google’s Visual Search on the Google’s Research web site: http://googleresearch.blogspot.ca/2013/06/improving-photo-search-step-across.html There is also a recent Wired story on Baidu’s visual search engine as well: http://www.wired.com/wiredenterprise/2013/06/baidu-virtual-search/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+wired%2Findex+(Wired%3A+Top+Stories) […]

G. Hinton wants to train Neural Nets with Trillions of Parameters

Focus of the recent Wired article was again deep learning researches being done at the Google X-labs. In that same article G. Hinton mentions that he would like to train Neural Nets with trillions of parameters:

“I’d quite like to explore neural nets that are a thousand times bigger than that,” Hinton says. “When you […]

UoT’s Deep Learning group’s startup acqui-hired by Google

Google acquired the new startup, DNNResearch, established by Geoffrey Hinton and his two graduate students Ilya Sutskever and Alex Krishevsky. The same team won the Imagenet Challenge in 2012. Hinton and his team is going to focus on improving the Deep Learning applications already being used by Google.

As part of his new job, G. […]