Deep Learning Program Simplifies Your Drawings | Two Minute Papers #107

Dear fellow scholar This is two minut papers with kly zsn fehér First let's

talk about rester and Vector Graphics What do These Terms Mean exactly A rest

image is a Grid Made Up of pixels And For each of these pixels we specify Color All There Is In An image It Is Nothing But A collection of pixels All photographs on your phone and generally most imes encounter imes res of More gds theer is more details however in Return If We disregard compression techniques the file size grows proportion to the Number of pixels and if we Zoom in two close We shall Witness These Classic staircase Effects That We like to call aliasing however If We are designing a website or a logo for a company Which should Look Sharp on All Possible devices and Zoom

levels Vector graphics useful Vector images are inherent from images as base elements of the image not pixels but vectors and control Points the difference is like storing The Shape of a Circle on a lot of pixels Point by Point Which Would Be A raster image or Just saying that I want to Circle On These coordinates with a given Radius and As you can see in this example the point of is to have raz Sharp images at Higher Zoom levels As unless sizes ofar small Vector imes Because We are not Colors Of Millions of pixels We Are Only storing Shapes If we want to Sound a bit more journalis we can Kind of say that Vector images have Infinite resolution we can Zoom in As much as We wish and we Won't Lose Any detail during this process Vector is process wey to

Convert image to a Vector Image

process image tracing imedi arises Why are we not using Vector Graphics everywhere well One the smoother the Color transitions And The More detail we have in our images the quicker the Advantage of Vector evat and two also Note that this procedure is not trivial and We are also Often at the Mercy of the Vector algorithm in terms of output quity It is Often unclear Advance inp now we need to able to understand andate Amazing piece of work the input is a Rough Sketch that is a rest image and the output is a simplified cleaned up and Vector version of It We are not Only doing Vector but slif as well This is a game Changer because This Way we can Lean On The addal input images sketches hand images und vorp Sketch simpli in each of these cases It is absolute Insanity How well it works Just Look at These results the next Question Is obviously How does this Wizard happen It happens by using a Classic Deep Learning Technique a convolution neural Network of that trained on large Number of input and output pairs ur Network diff archit ased series of up sampling convolution steps intui This algorithm learns a sparse and concise representation of these input sketches This Means that it focuses on the most defining features and throws Away all the unne fluff and the up sampling convolution step syi High resolution images That We Can Easily Vector using Standard algorithms It is fully automatic and requires no user intervention in case You Are scratching Your Head about These convolution we have had plenty of discussions About This peculiar term before I have linked the appropriate episodes in the video description box I think youl find Them a lot of Fun in one of them I pulled Out a Guitar And added reverberation To It using convolution It is Clear that There is a Ton of untapped potential in using convolution variations in Deep neural Networks we have seen in a Deep Mind paper earlier that used dilated convolution for State Of The Art speech synthesis that is a novel convolution variant and this piece of work is no Exception either There's also a cool online demo of this Technique that anyone can Try Make sure to post your results in the comment section would love to have a look at your findings also this two minut papers fan Art a nice Little logo One Of Our kind fellows Sent in really Great to see that you taken your time to help out the series very Kind of you thank you Thanks for watching And For Your generous support see you next [Muziek] time