Robots learn to cook by watching YouTube

[Rohith_Kumar_Sp]

When it comes to learning how to cook, it turns out that robots may not be so different from humans after all... or are they?
 

When it comes to teaching robots how to do things, there are some very key differences. A human knows what you mean when you say "I need a cup". A robot needs to be taught that that means it has to turn around, go to the cupboard, open it, take out the cup, close the cupboard, turn back around, return to you, manoeuvre the cup over the bench, and release the cup.
 

This is one of the key parts of figuring out machine learning: How can you program a robot so that it can intuit that a plastic cup, a glass and a mug may all be classified under the general term "cup"? How can you design a robot that is able to teach itself?
 

One way, as researchers at the University of Maryland Institute for Advanced Computer Studies are finding out, is YouTube. More specifically, cooking tutorials on YouTube. By watching these videos, robots are able to learn the complicated series of grasping and manipulation motions required for cooking by observing what humans do on the Internet.
 

"We chose cooking videos because everyone has done it," said UMD professor of computer science and director of the UMIACS Computer Vision Lab Yiannis Aloimonos. "But cooking is complex in terms of manipulation, the steps involved and the tools you use. If you want to cut a cucumber, for example, you need to grab the knife, move it into place, make the cut and observe the results to make sure you did them properly."
 

The robot uses several key systems in order to learn from YouTube videos. Computer vision, with two different recognition systems, allows the robot to visually process how the presenter grabs something, artificial intelligence processes that information, and finally language parsing helps it understand spoken commands and translate it into an action.
 

In this way, the robot can gather individual steps from various videos and assign them rules according to its programming, putting them together in the correct order.
 

"We are trying to create a technology so that robots eventually can interact with humans," said UMIACS associate research scientist Cornelia Fermüller. "For that, we need tools so that the robots can pick up a human's actions and track them in real time. We are interested in understanding all of these components. How is an action performed by humans? How is it perceived by humans? What are the cognitive processes behind it?"
 

The difference, the team said, between their research and previous efforts is that they are concentrating on the goal, not the steps. The robot can draw on their databank of actions to string them together achieve the goal, rather than copying verbatim, step-by-step, a series of actions.
 

It's a system that apparently works. According to the team's paper, the grasping recognition module had an average precision of 77 percent, and an average recall rate of 76 percent. For the object recognition module, the robot achieved an average precision of 93 percent and an average recall of 93 percent. Overall, recognition accuracy on objects was 79 percent, grasping was 91 percent, and predicted actions was 83 percent. The drop in object recognition accuracy was because the robot had not been trained on some objects, such as tofu.
 

"By having flexible robots, we're contributing to the next phase of automation. This will be the next industrial revolution," said Aloimonos. "We will have smart manufacturing environments and completely automated warehouses. It would be great to use autonomous robots for dangerous work -- to defuse bombs and clean up nuclear disasters such as the Fukushima event. We have demonstrated that it is possible for humanoid robots to do our human jobs."

The team will be presenting their research at the Association for the Advancement of Artificial Intelligence Conference in Austin, Texas, on January 29, 2015.

https://www.youtube.com/watch?v=xRwhsFtUnpc

Source : http://www.cnet.com/news/robots-learning-to-cook-by-watching-youtube-videos/

[GhostWindGamer]

OH DEAR GOD THEY ARE LEARNING

[Opcode]

[ShadowCaptain]

thats how I learned to cook

 

AM I A ROBOT?

[soaringchicken]

OH DEAR GOD THEY ARE LEARNING

PLS NO Skynet

[eelnico09]

It's coming....

 

"Robot learns how to search on youtube"

"Robot watches video about murder"

"Robot learns how to kill"

[GhostWindGamer]

Spicy meatball

Thats a spicy way to get killed by robots 

[Rohith_Kumar_Sp]

thats how I learned to cook

 

AM I A ROBOT?

Who said anything abt being a robot if you learn from the internet  

[GhostWindGamer]

It's coming....

 

"Robot learns how to search on youtube"

"Robot watches video about murder"

"Robot learns how to kill"

This is my biggest fear 

[Rohith_Kumar_Sp]

This is my biggest fear 

it's only as the maker created it , Robots have no gain or motive to kill us , Humans may be greedy and jealous to kill an other person but a robot isn't 

[Opcode]

PLS NO Skynet

Not anytime soon. A robot doesn't have the "brain capacity" to handle enough information to go beyond basic tasks.

[soaringchicken]

Not anytime soon. A robot doesn't have the "brain capacity" to handle enough information to go beyond basic tasks.

It's all in good fun man, I'm kidding

[ShadowCaptain]

Who said anything abt being a robot if you learn from the internet  

 

clearly I was joking.....

[Rohith_Kumar_Sp]

clearly I was joking.....

so was i ,well atleast i meant it that way ,English is not my first language , pardon me .

[GhostWindGamer]

it's only as the make created it , Robots have no gain or motive to kill us , Humans may be greedy and jealous to kill an other person but a robot isn't 

You say this now

[eelnico09]

Not anytime soon. A robot doesn't have the "brain capacity" to handle enough information to go beyond basic tasks.

Idk about you, but i have at least 40 more years of life, now think how everything was 40 years ago(1975) and how's now, and imagine how shit's going to be in 40 years.

 

This shit is not that distant.

 

 

[ShadowCaptain]

so was i ,well atleast i meant it that way ,English is not my first language , pardon me .

 

hehe ok

 

 

it's only as the maker created it , Robots have no gain or motive to kill us , Humans may be greedy and jealous to kill an other person but a robot isn't 

 

 

If robots gain intelligence, they might realise we are destroying the planet, and causing wars etc, maybe they will kill us because we are bad for the planet

[AlwaysFSX]

I would love a robot that can cook for me when I'm lazy. In a closed environment though to prevent that. V

 

It's coming....

 

"Robot learns how to search on youtube"

"Robot watches video about murder"

"Robot learns how to kill"

*flings kitchen knife*

[PercySavageEG]

First the Japanese overlords and their vehicles of destruction, now the impending mayhem of pan-wielding learning death machines? Whats next? IRobot and Skynet shit?

[Opcode]

It's all in good fun man, I'm kidding

 

Idk about you, but i have at least 40 more years of life, now think how everything was 40 years ago(1975) and how's now, and imagine how shit's going to be in 40 years.

 

This shit is not that distant.

 

 

-snip-

I personally feel as if we do harness the technology to be capable of at least building a semi-retarded terminator.

 

[AlwaysFSX]

I personally feel as if we do harness the technology to be capable of at least building a semi-retarded terminator.

Convince it that bouncy balls are really dangerous and it having the intent to kill.

[Rohith_Kumar_Sp]

If robots gain intelligence, they might realise we are destroying the planet, and causing wars etc, maybe they will kill us because we are bad for the planet

The first thing you do when you develop an AI that is self aware is to ask it to create  Robotic laws that can not be broken , namely don't not kill humans and nor can it hurt anyone laws , it will write the code, shut it down , Create another AI implement the law, Robot can only understand the value of saving earth if it is the primary programming to save earth , it isn't , i has no purpose to kill humans to save earth , save earth and do what exactly ? you get what i mean ? 

[ShadowCaptain]

The first thing you do when you develop an AI that is self aware is to ask it to create  Robotic laws that can not be broken , namely don't not kill humans and nor can it hurt anyone laws , it will write the code, shut it down , Create another AI implement the law, Robot can only understand the value of saving earth if it is the primary programming to save earth , it isn't , i has no purpose to kill humans to save earth , save earth and do what exactly ? you get what i mean ? 

 

If it had TRUE AI, then it would be able to break rules and it would understand the concept of self gain

[Rohith_Kumar_Sp]

If it had TRUE AI, then it would be able to break rules and it would understand the concept of self gain

you should see a movie called Automata , it's really good 

[ShadowCaptain]

you should see a movie called Automata , it's really good 

 

Is that the one with Antonio Banderas? looks interesting, will definitely watch it