A.I. Fighter

[EZbakeROFLcakes]

You could try making advance AIs in UDK :D

[xistoxs]

Thats awesome. I hope that NPCs become super intelligent really soon.

[TheReasonWhyGuy]

you should have 2 train against each other...

[CyberWalker2004]

Really nice background music!!!

[Houshalter]

But how does it know what action is best in a given situation? In order to train it, you have to have an example of the best possible answer in every situation.

[virgildog1950]

@Houshalter Keeping your health up and hurting the player reinforces the bot's learning because it complies with its objective.

[rcparts]

@virgildog1950 It could have been done the way you said (like TD-Gammon did), but they've used a different approach. It's on their site.

[dajwilkinson]

@Houshalter Maybe every time AI loses, its last move (or set of last few moves) is the basis for the back-propagated error signal? Simple game-learning by this technique has been around for decades - e.g. Martin Gardner, Further Mathematical Diversions (1969), chapter "A Matchbox Game-Learning Machine".

[Ambakapu]

I like your story telling method .)) fun! great job!

[trisky1234]

right, now code a FSM for a bunch of basic 'emotional' states the AI entity can be in, e.g. pain, fear, bravery, victory, etc, and implement a neural network underneath to handle the players moves accordingly -> evade when in pain or fearful due to player speed, attack when feeling brave or victorious due to players slow movement :)

[Dodekus]

Epic video. I'm thinking of doing a neural network for a game. Try at least.

[scop3nz]

+1 Well done.

[clustro]

this video is beyond epic

its WICKED SICK!

[fidonoso]

excelente video.......

Genial el futuro de la I.A

[Illithan]

Nice way to train the AI player against any particular human. I suppose in practice you could use a default neural network to begin and train using pseudo-rehearsal to strengthen particular input/output rules whilst interacting with the player. But If there is a winning strategy in this game then the NN may easily find it.

[Skinnyram]

Did you use an engine that is avaliable to the public? I'm wanting to do Neural networks for my final project but having problems locating some code thats relevant.

[skeelogy]

Nope, it wasn't done using any engine. Some basic source codes can be found in neural network books I believe, so you can try digging into those.

[drowjam]

The biggest problem i have noticed with neureal nets is that they can be trained to do on thing but if you train them to do another they will often not be able to complete the orginal task they were assigned. I had an idea for an AI i once that involved using neural networks to search relatioanal databases to reconize patterens in a players ablility but that was just a little to much for me. anyway great vid.

[Conscars]

I don't understand how the weights for the hidden layer inputs are calculated. How are we supposed to know the error for each of these hidden units?

[KaarsUnderlost]

just as you would train a 'live' creature.. if you want to train it into something you know what the outcome is supposed to be.

[Althar93]

Cool, just what I needed : an explained implementation of neural networks!!

You can find plenty of stuff on networks, but they never talk about a specific implementation, rendering it impossible to fully understand.

Thanks for the vid, and awesome work!!

[fopple13]

great vid, im trying to make a robot using neural networks, it makes a lot more sense now!

[blenderuser]

Great video! I really needed an example like this, as I was a bit confused about what the inputs and outputs could represent. Cooool

[DrJBN]

Great vid & idea. Backprop is prone to "catastrophic interference" Does your player have trouble initially retaining old skills when learning new ones?

[skeelogy]

Yes it does actually. It'll tend to forget some of the combos but that's kind of expected with the interference from other samples, like what you have mentioned. Seems like you have done quite a bit of research on this topic as well :)

[jopir]

Good explanation my child

[msj120]

Nice Video, well done.

[adisharr]

Very interesting example :)

[lanflett]

Impressive and very didactical video. Congrats and keep up with the excellent work.

[ppaaii1111]

That's damn good

[rcparts]

Congrats! Very didactic and funny movie! I already know ANN's well, but if I didn't know it, sure I'd learn it with your video =) (at least the basic mechanism and the motivations)

[dalyk]

Awesum!

[A3A3adamsan]

Fun and educational, congratulations! :)

[mpeniak]

brilliant!!!GREAT WORK

[DR3TH]

thats why i always lose at tekken.

[sd19matt]

By the time the AI had learned his combos....the computer player would have been defeated. Neural Networks can never be perceived as perfect as human perception.

[red15nls]

Hmm, the time it takes us humans to figure out how to do things takes 10 to 20 years and sometimes even longer. Which we call growing up :) So the AI needs to grow up too. Just like you need to learn a new game the ai has to learn your abilities and how to react to them.

[GBart]

Our brains are nothing but really complex neural networks.

[8488463p0wnd]

This video is good but I still don't understand very well. Does anyone know a good place to learn more?

[einsena]

if you aren't a teacher, you should be lol!

[danuthaiduc]

OMG now i understand neural networks!!!!!1

[UlisesyAfrodita]

¡Buena idea! Una forma muy divertida de experimentar con redes neuronales

[jabbasaw]

Great work guys. It provides a really interesting contrast between rule based systems verse neural networking.

[nibble222]

well done! interesting movie :)

[kirbyfan95]

sweet, i like the story of how you taught the AI.

[futurehorizon24]

From the way you described it this game is no Neural Networks but just simple if statements. Dont post things like this when there are people actually trying to develop real methods.

[skeelogy]

It's kinda strange that you wrote that there's no neural networks involved. Have you watched the whole video? We have another version of the game which is running on Finite State Machine that uses if-statements. But that is used as a contrast and is not the feature of this game. Perhaps you can watch the later part of the video which talks about the neural networks.

[sevketcakir]

Hi. I wonder how many human players trained the network? Fighting styles may differ from person to person. Is the A.I. fighter successful against another player who didn't train the network?

[skeelogy]

Actually we haven't had the chance to do extensive tests on it yet because we are busy with another project. We will probably release it to the gaming community some time later so that they can help us to train the network.

[Fromzon]

does this game really exist, and if so, where can i find it?

[skeelogy]

Yes this game exists. The game fighting parts above is done by capturing videos of the actual game itself. Unfortunately the game cannot be released yet due to some issues but we do have plans to release it some time this year.

[frogger626]

Excellent! I can't wait for this thing to come out!