Autonomous Checkout, Real Time System v0.21

Deep learning, meet brick and mortar.
This is a real time demonstration of our autonomous checkout system, running at 30 FPS. This system includes our models for person detection, entity tracking, item detection, item classification, ownership resolution, action analysis, and shopper inventory analysis, all working together to visualize which person has what item in real time.
We're excited to demonstrate more features for our zero friction checkout experience. Stay tuned to see the future of retail.
www.standardcog...

[@setme31]

This was a great video and exactly what I was looking for. I can't believe the smarts of this system. It even recognized when you were passing items to each other.

[@ChristinaMagma]

This is actually impressive, and i bet the technology has improved so much since then

[@JorgeGamaliel]

Awesome!! Neural Networks, deep learning and applications in surveillance.

[@achonwachris8431]

Amazing, simply outstanding.. Been working on something like this, since you guys found the solution first.. Am so so amazed.

[@mubasharnisar7328]

Could you please share research papers related to this system.
If you have dataset please do share with me.

[@Albinorama]

the end of shoplifting....I will have to find job now.

[@george_tino]

Great! I have always being dreaming of this. Thanks to the neural team who developed this up. I just realized Amazon is making something closely similar.

[@sudeeppendyala5949]

Brilliant System...,just loved the work you've done.

[@seagoat666]

Awesome! cant wait for it to be every where; that can make it possible for 24/7 stores

[@honestreview9019]

we are watching you - its amazing

[@MrLezraf]

Is this system able to tell the difference between differents types of cookies for example ? In this video, it seems that you have only one type or brand of each product.

[@Nick-zu9sn]

Hey maybe it's just me, but I think having one of the guys be a juggler and toss a few juggles in there with 3 different items would have spiced this up a bit.

[@arifazizi4869]

could you tell me please, what technology you'r use?. its look cool and fun, thumbs up for your creation

[@haviettien1914]

Quick question, every store has different products. How do you recognize all of them. Do you have to customize the object recognition model for every shop?

[@lokilance]

It looks great however I have a question. Can it keep up with say holiday shopping? For example when you get say around 20 people in the same small area.

[@MrRaianknight]

Question is really important. Do you have answer for it?

[@siglange4882]

I can imagine they limit the number of people allowed in the store at a time.

[@viljar1112]

you gotta use logic _ if you can't see the the picture then you can't detect the taken object, so yes some one could easily steal turning hollidays mess. However when finaly people starts to clear up and camera can take the new picture it can see and calculate the missing items versus purchesed items. Then it can pin point the exact moment when and probably who took it . ( I bet facial recotnicion system will also be included). But as a honest costumer you should probably show the camera whot you took befour you put it in the bascet. Markets can also install more cameras = more view angles = more relayability.

[@villa-tech9058]

the fail safe will be technologies like RFID tags that can be tag on each item to cover where CV was not able to track

[@Jagerleaf]

Sig Lange therefore limiting business which would cause a decrease in sales; makes you wonder why even implement into a store if you’re limiting your customers

[@ChengliangShan]

Awesome!!!

[@florianwilk5387]

Great work guys!
How many cameras did you use / are needed at least for this scenario?

[@sazidabintaislam2676]

Are you using Deep CNN for this detection method? Where the computation is going on, in the computer?

[@arifazizi4869]

wow. awesome!!

[@DanieleGiorgino]

Can it recognize someone trying to steal using the cooler?

[@chienyu6056]

Is it recognizing the hard-coded locations of each item, or is it actually recognizing the image of what the item is?

[@sean092885]

there is a part where they grab misplaced items from the wrong shelf. unless it was setup to be "misplaced" in a certain place... i'm fairly certain its identifying the item. not to mention walking off screen and back on with items still recognized.

[@mehmetfurkanatilgan]

@@sean092885 If cameras really identify the items then thats awesome..

[@richardbarrera1229]

Do you where can i buy this, is for a project a small store for my hotel.

[@unsa8000]

super, mais la question Loki Lance est tres actuelle!!! mais on a pas la reponse

[@billyAgnaska]

Any chance I can get similar code to this system?

[@madbull2360]

Am i the only one thinking of hiding the camera’s view with a tape will work ? 😂

[@fmindorderflow7270]

it is very easy to fool AI if you know a bit about image detection :)

[@mubasharnisar7328]

I am working on the same project, anyone who can suggest me some research papers related to this.
Or anyone have dataset.
Please do share with me.

[@kanhataak1269]

Hello mubashar are still working on it.

[@Willyliemfams]

where can i buy this software?

[@younglam8470]

I.like....

[@ЧайкаСергій]

Прікол

[@anwarhmdn]

Amazon go haha

[@jameswalker3980]

All clowns making comments below.