Evolution is rarely linear. As with other technologies before it, deep learning has followed a series of step functions defined by sudden, often unexpected, outbreaks of capability
Notion hosts regular, free ‘Enterprise Innovation Briefings’ where we invite 30 C-level people from large enterprises, giving them an insight into the world of emerging technology and venture capital. Our next briefing will address Computer Vision, including experts on Artificial Intelligence (AI) and Machine Learning.
Ahead of our briefing on February 7th, Dr Christian Thurau, CBDO and Co-founder of Twenty Billion Neurons GmbH gives us a sneak peek from his talk, discussing how, in future, computers will learn common sense world knowledge through video.
Christian has over 15 years of experience in data mining, machine learning and pattern recognition and holds a PhD in data mining and machine learning from Bielefeld University, Germany.
“Each step function fundamentally pushed the envelope beyond what computers were previously able to achieve. One of the first such breakthroughs came in 2012 when Alex Krizhevsky, Ilya Sutskever and Geoffrey Hinton showed that deep neural networks, trained using backpropagation, could beat state-of-the-art systems in image recognition.
Since then, similar breakthroughs have occurred in previously intractable problems.
These range from machine translation and voice synthesis to beating the world champion in the game of Go. Importantly, each milestone seemed out of reach at the time, which made their achievement even more surprising.
At TwentyBN, we believe that the next breakthrough in Computer Vision will concern video understanding.
The vast majority of commercial use cases in video understanding are stretching the envelope of what current technology is capable of.
The reason is, that machines today are missing a common-sense understanding of the real world. We humans rely constantly on our innate ability to understand and make inferences about our environment as we navigate the physical world.
Our common-sense reasoning is built up through lifelong experience, beginning in early childhood and extending well into adulthood.
Over the past year, we have created spatio-temporal video models, video infrastructure, as well as a data operation that allowed us to create millions of labeled videos, showing everyday common-sense scenes and situations – many of them designed to be extremely subtle and hard to distinguish.
This allowed us to successfully train neural networks end-to-end on a wide range of action understanding tasks, that neither hand-engineering nor neural networks had appeared anywhere near solving just a few months ago.
These recognition tasks not only drive commercial value at TwentyBN but also drive our long-term AI agenda, which represents another, longer term, bet on learning common sense world knowledge through video.”
Our next event is being held on February 7th morning/lunch at Notion’s London office (Marylebone) and the theme is Computer Vision.
We are delighted to host as a speaker Prof. Philip Torr from Oxford University, who is an award-winning world expert in computer vision systems. Prof. Torr is also an advisor to 5AI, one of our portfolio companies and will be presenting along with Stan Boland, British tech guru and 5AI’s CEO.
• Date: Wednesday, 7th of Feb 2018
• Time: 11am-4pm
• Venue: 91 Wimpole St, Marylebone, London W1G 0EF
To register your interest and join the waiting list, please click here