Artificial Intelligence Conference 2018 – London, United Kingdom

Sobre este curso

if you have small kids you may know Wally this is a Pixar movie from only ten years ago and this little guy basically his job is to cruise around a post utopian world and clean up the

trash and he has been doing this for seven hundred years on solar power and the way he does this is through computer vision technology he has the ability to identify things that are trash and not

trash and sort them appropriately put him into his chest and compact them but it’s only been ten years and you know this is no longer that far off I mean here’s a customer of ours hikvision

which is a leading AI company in China and this Factory that they’ve automated with these little robots ships about 15 million dollars worth of product every single day and by automating the entire

sorting process they’ve reduced their cost by 80% and at the same time they’ve increased the efficiency of this particular warehouse by 60% there used to be just hundreds of workers in this

building and they completely retrofitted it to enable these cute little robots that are operating entirely autonomously using computer vision optical sensors on the robot to help them determine where

they are and understanding where they need to be pretty cool here’s another example The Terminator the t800 whose primary directive is to destroy humanity by finding Sarah Connor and eliminating

her before she can give birth to her son also a user and fan of computer vision technology what the Terminator also has is natural language processing so he has the ability to communicate and he’s got

a neural net processor CPU which enables him to be a fully functional cybernetic organism what some of our customers are doing rather than destroy humanity they’re trying to save humanity by

seeking out to destroy diseases in this case GE healthcare is using and production to be able to understand highly complex CT MRI PET scan sonogram mammogram images and what’s really

important is that you know these types of images can’t be compressed they really have to be looking at the full res image in order to get the depth and the complexity of the image that they’re

looking at in the case of GE they leveraged a tool kit from Intel along with xeon processors and they’re now getting 16 times better performance on intel architecture than they were

getting prior and pretty interesting case study that has an impact on everyone’s lives and of course no conversation of droids would be complete without c-3po and r2d2 c-3po also

computer vision user fluent in six million forms of communication and also uses machine learning to detect people’s emotions in the case of r2d2 an astromech droid used to transport secret

rebel communications as well as repair starships operates in low-power mode very important in fact what you’re seeing here is a lot of these droids because they’re operating outside of a

data center they do require low-power and they do require high degrees of computational capabilities to handle all of the inference that’s coming in through the variety of sensors that they

have and so what we see is that vision is becoming almost the ultimate IOT sensor and it’s transforming every industry so in retail we’re using computer vision technology to allow for

staff ‘less stores so now in the u.s. it’s Amazon go that’s rolling these out in China it’s JD comm and now shoppers have basically the ability to replicate the online experience offline which

means online it’s a one-click purchase customized and personalized for your individual needs in the case of a staff ‘less store you walk in you take whatever you want off the shelf and you

walk out there is no check process we’re gonna talk about a couple of industrial cases in transportation computer vision technology is going into vehicles with gaze control so you can

understand if the driver is distracted and looking at their phone or if they’re in fact dozing off hugely important for safety in smart cities digital security and surveillance has been the key

workload or use case for computer vision technology particularly here in London which led the way with a completely blanketed security system across the city that allows city operators to

actually track people and vehicles from camera to camera in a networked way but it’s gone way beyond just security and surveillance now cameras are being used to help automate traffic situations to

automate the parking situation and really help with pedestrian movement and a number of other things