Deep Learning Applications Deep Learning Applications In Real Life Deep learning Simplilearn

Author:

Simplilearn

Keywords:

deep learning applications,applications of deep learning,deep learning applications in real life,cool deep learning applications,deep learning example,deep learning medical applications,deep learning applications in healthcare,deep learning in image processing,deep learning in music,deep learning in composing music,deep learning in robotics,deep learning applications 2019,deep learning,deep learning simplilearn,simplilearn

Subtitles:
deep learning applications by simply learn my name is Richard Kirchner with the simply learned team that's WWLP learn comm get certified get ahead and today's exciting world of AI artificial intelligence we're gonna cover some of the many uses is being applied to healthcare entertainment composing music image coloring robotics image captioning advertising earthquake prediction and there's many more it's just an exciting time to be in the field of AI and deep learning and artificial intelligence now all these fields are just booming it's just amazing what's going on so let's take a quick glance at some of the things going on and we'll start with our healthcare we take a look at healthcare deep learning is reshaping healthcare industry by delivering new possibilities to improve people's lives and we have computer aided disease detection I know a number of startups have come up where they're working on you take a picture like if you have some kind of irritation or rash and they use that or even you know some infection in the eye and they use the deep learning to help identify what it is hey this is 99.99% nothing don't worry about it or this is 93% chance and usually doesn't come out ninety three percent chance if you do deep learning but it says hey you should probably take this into the doctor you know I'm take a look at it and see what it is these are the things that could be analyzing genomes that's a big one very controversial nowadays with all the different genome editing and options they have and things are doing but deep learning is definitely diving into the genome projects we have discovering new drugs and this there's so many directions for new drugs going on it's like one of the booming industries especially in the US but all across the world and they range from exploring different plants and the opening up of let's say marijuana in the US and number cities are exploring that for medical use - how do you reprogram T cells in the human body to combat disease there's whole industries and data analysis another one is chemistry how do you use a chemical cell or a chemistry molecule not a chemical molecule to imitate the T cell so you could then put those molecules in there and then they grab the diseased cells out or the cancer cells or whatever out just like the human body would so discovering new drugs and how to apply that is just a huge industry all these are huge industries and medical imaging will even take a glance closer at that but being able to analyze you know we get your MRI you get your cat scans the doctors can spend a long time looking at those and taking careful measurements and most of that's done by hand but as we start applying deep learning that deep learning can do a lot of that work not only do it the same but it be uniform from one hospital to the next meaning that as a industry it continues to progress and you get better results and better services so deep learning is reshaping healthcare industry by delivering new possibilities to improve people's life and deep learning helps an early detection of cancer cells and tumors improves a time-consuming process of synthesizing new drugs and inverse sophisticated medical instruments and not even sophisticated they're using iPhones in deep learning to help with medical diagnosis so it's at all levels it's just amazing what they're doing in the healthcare industry with deep learning and they've just barely tapped what there's possible they're just now evolving into the beginning of this where it's exploding out there so entertainment that's always a fun one a lot of people go into the entertainment and deep learning because it's just fun you know what's my favorite movie all that fun stuff deep learning is used in the entertainment industry such as Netflix Amazon and filmmaking and you have like your recommendations that's probably the most common one you see and you can see here where they have India transform the true story the third pillar now markets in the state leave so it comes up and says hey this is based on what whoever and of course this is showing up on the Amazon or Amazon Prime Netflix does a good job of grouping movies for you it says since you like these movies you might like these movies over here same thing with VEVO it does that also Amazon Netflix and VEVO use recommender systems to provide personalized experience to its viewers using their show preferences time of access history etc so really this is one of the cool things is instead of having to dig for the information deep learning starts doing that for you so instead of spending hours trying to figure out what you watch you can actually just enjoy the movies and have them come up deep learning is used in the entertainment industry such as Netflix Amazon and filmmaking continue on and deep learning we have things like the IBM Watson this is just really cool I didn't know this till I read this till we did this slide wimbledon 2018 used IBM watson to analyze player emotions and expressions through hundreds of hours of footage and then it auto generated highlights for the telecast so you can imagine somebody sitting there looking through all the boring stuff I want to look at the interesting stuff it helps us sort that out again you know just like the Netflix and referral it helps you find your choices without having to do all the heavy work of digging through really bad berated movies and music and audio generation another huge part of the industry and this kind of blends right in with video and moviemaking voice and audio recognition technology can be used to train deep learning network to produce music compositions so it learns the patterns and generates new music Gogol's wavenet and betas deep speech can train a computer to learn the patterns and the statistics that are unique to the music it can then generate a complete new composition that's a little scary I always thought that creativity would always be the human factor that we'd always have our creativity but you know computers can do some of that too and then we have our reads lip movements oxford and google scientists created a neural network called lip net that could read people's lips with 93 percent success this can be used to add sounds to silent movies I like that it'd be fun to hear what they actually were saying back there they probably said something totally different than what's actually on the movie that'd be pretty funny on the scary note this would allow surveillance equipment to pull what people are saying even if they're really far away as long as they can see the lips that's a little scary coloring images and videos deep learning can be used to color images and videos by taking the objects in their context within the photograph so you have your black and white image is fed in and you have a convolutional neural network is used to learn the patterns and usually you train it you train it with hundreds of black and white photos that you already have the color image for us you know to expect and they take the new one and it recreates the image by adding the colors and this is also used in the movie industry to add in special effects there's all kinds of cool things they can do with this that goes way beyond just coloring pictures of course the coloring pictures is something that you can probably do in your own lab on your own computer and build your own neural network to do that robotics one of my favorites robotics all these tools that make them so exciting is that you can go out and even now used to be you'd have to spend four five six ten thousand US dollars to buy your own robot and then start programming it and doing neural networks now it's around two thousand dollars and that continues to plummet there's even kid robots much cheaper robotic hands all of that that you can purchase and generate your own neural network and start trying to figure out how to get it to do things and so we talked about deep learning it's heavily used these days in building robots to perform human-like tasks robots are built to understand the world around them very important how to figure out what's what I remember if you go back 20 years they could not get a robot to figure out the difference between a coke can and a pencil because they're both cylinders and they just they just had no idea how to do that back then you know they did the beginning of robotics really the neural networks was mid 90s where it started really taking off and the theory started coming in and so you go back to that time and figuring out those differences was almost impossible before them and they trained to make better and safer decisions without supervision and you can see in this case you know in outer space having a robot up there you know if the window blows out air pressure is lost the robot is not going to be hurt by it they don't need air and oxygen and their decisions usually instead of being like a human's tired in this case the robot never is tired so you know there's definitely a lot of tasks for a robot that can be used replace humans would be safer think nuclear reactors have melted down after the tsunami in Japan using robotics to go in there and clean up would be really important instead of putting a person to hazmat suit that my or might not protect him from the radiation in Boston Dynamics we have spot mini he's fun to watch we were Pope the video or the Atlas robots by Boston Dynamics react to people pushing them around get up when they fall can unload a dishwasher and do other tasks as well so we're starting to see this kind of robotics coming in to the house the Atlas obviously is a very expensive robot but there is certainly more common ones you can actually bring into the house and then spot minis fun to watch run you can I forget how fast he goes because pretty fast image caption generation another ended part of the industry in deep learning deep learning is used to display the contents of an image I love this with my google photos the system is trained with very large convolutional neural networks for detecting objects in the photographs and then a recurring neural network like an LS TM is used to turn the labels into coherent sentence dog catching a ball so this is awesome this is like you go in your google photos and you say hey I want all the pictures of my dog or I want all the pictures of kids or whatever it is I'm thinking to the point we can now do I want pictures where my dog is catching the ball advertising one of the better paid fields in AI is in marketing and advertising especially if you are working on bonuses and commissions so advertising an advertising deep learning allows optimizing a user's experience so deep learning helps publishers and advertisers to increase the relevancy of the ads and boost the advertising campaigns allows a network to reduce costs by dropping the cost per acquisition of a campaign from 60 to $30 create data-driven predictive advertising real-time bidding of their ads and target display advertising and again this goes from experimental where I'm just gonna guess what people want and will put out a pink sign that says you want to add on widget to your order well nobody wants widgets and nobody wants to see the pink sign maybe you got three people want to see the pink sign and you find out that the blue sign a lot of people are reading that and they really want a widget number two whatever that is so when you talk about advertising it also deals with marketing and also preference of merchandise so becomes really huge as far as what your advertising and generating profit and sells for a company and that's across the board whether you're talking big real estate or property management and rentals or Amazon uses this for all their merchandise coming out of their warehouses any of those are generated from the AI generates more sales and that's the bottom dollar business is always looking at the bottom dollar where can I make money on this earthquake prediction so now we're getting to something safety human safety you also think weather predictions all those fun things deep learning model can be used to predict earthquakes by considering a factor called von Mises yield criterion scientists at Harvard use deep learning to teach a computer perform viscoelastic computations that is quite the mouthful which are used in prediction of earthquakes this application helped to improve the earthquake calculation by fifty thousand percent that's amazing we went from just guessing when an earthquake is gonna happen where we have almost no reliability to being able to have a decent guess as when the earthquakes are gonna hit that's pretty impressive so that was a brief tour of all the cool things going on in AI definitely don't let those stop your imagination from thinking about what other tools can be used with AI to generate but you can also look at these as major ones if you're looking for work and you're saying hey what do I enjoy doing do I want to go into entertainment do I want to work in the medical industry which is very lucrative entertainment can be lucrative but it's very hard to get to that level robotics robotics is exploding no one knows where it's gonna go I was just reading an article on mini tensor flow for embedding deep learning models once you've taught them onto a cheap processor so now you can use that to do some very basic simple things wherever you go with your understanding of deep learning and the AI field I want to thank you for joining us today and I wish you lots of luck for more information visit www.gfi.com/webmonitor hi there if you like this video subscribe to the simple learn YouTube channel and click here to watch similar videos to nerd up and get certified click here

Loading