Cool , but when are they starting to make useful humanoid robots to do digging in mines, clean streets ,carry stuff around, replace fireman etc ... So that we can get some free time and maybe play pool instead of the robot :)
@Braniti muy cierto lo que decís, pero como lo vamos a lograr si una respuesta coherente como la tuya tiene 5 votos y alguien que observo una falta en el minuto 2:14 tiene mas de 100 votos.
- the stroke difficulty is only calculated based on the angular deviation of the white ball, meaning how much in degrees one can aim to the left or the right still being able to pocket the object ball. This measurement implicit includes the distance white ball - object ball, object ball - hole and the angle to pocket. No weighting of factors is necessary.
- cushion shots might be possible as soon as we get a better pool table ;-)
This is amazing pioneering, and to think it wont take many decades to refine this technology to the point it is aesthetically pleasing, and integrated as a function into humanoid robots without the need for an overhead camera. It would give as a companion to play pool against. However little we may win :P
@OwnageEngine but those factors need to calibrated: what's more influential, distance or angle? In what ratio? How much is the difficulty adjusted by having a ball kissing the white?
That data needs to be gathered by experimentation.
If they gathered it by observing humans then they could possibly claim that it shows the robot's ability is comparable to that of a human's in regard to potting difficulty, but they'd need to present a lot of data comparing the robot to humans to make that claim.
@Isanion Its arbitrary. The stroke difficulty would be related to how inaccurate parts of the robot are. If it has trouble aiming at balls at the other side of the table, but can hit close ones accurately then distance to the ball would have more weighting. So something like D = k(Distance to ball) + m(Distance to pot) + n(sin(angle to pocket)) where k,m and n give some indiacation of how bad the robot is at certain things.
@Isanion How is this pointless? Its exactly the same way you measure difficulity in anything humans do. If we arent good at something, its more difficult for us. Same thing happens with the robot.
@Isanion No, he has an algorithm to calculate the stroke difficulty for every shot. He was showing that the reuslts of that algorithm correlate linearly with the actual success rate, which is evidence that the algorithm works. He uses this algorithm during play so the robot can choose the easiest shot.
@Isanion The stroke difficulty is determined using the "allowed angular difficulty", which is basically the stroke's "error" angle that allows the object ball to be pocketed. The smaller the angle is, the easier the stroke will be.
Один вопрос - зачем это надо?
paraderecords 2 weeks ago
Hmm. Wonder what other nonsense are coming out of pool/billiards world already. Lols.
diconder 1 month ago
aimbot~!
coregunner55 1 month ago
Fucking ninjas!
YuriHisokaa 4 months ago
Why?
nykaenenable 5 months ago 2
Yeah, but can it buy you a shot and rack?
ireallyliketoparty 6 months ago
quite nice.
a robot can be VERY good at pool : )
DanFrederiksen 6 months ago
teach it how to bar fight then ill be impressed
spclizdfuse1 7 months ago 13
That thing would beat the shit out of me.
JohnBlack666id 8 months ago
and next... the revelion of the machines...!
jjpl2001email 8 months ago
So terminators will play Pool...thats lame...
monishfiji 8 months ago 4
@monishfiji with human skulls.
egokick 8 months ago
cool
catbomber24 8 months ago
Fascinating stuff, even though all I saw when I clicked was ....Armed Robot.
rigragtag 8 months ago 4
very clever
Skynet will just hang around in bars instead of taking over the world
fridgelawless 8 months ago 3
Cool , but when are they starting to make useful humanoid robots to do digging in mines, clean streets ,carry stuff around, replace fireman etc ... So that we can get some free time and maybe play pool instead of the robot :)
Braniti 8 months ago 7
@Braniti amen
123OrioN321 8 months ago
This has been flagged as spam show
@Braniti Well, we gotta start somewhere, right?
SuperMarioRPG6400 8 months ago
@Braniti muy cierto lo que decís, pero como lo vamos a lograr si una respuesta coherente como la tuya tiene 5 votos y alguien que observo una falta en el minuto 2:14 tiene mas de 100 votos.
hervann 7 months ago
Software fail, robot touch a ball with his right hand on 2:15
naxein 8 months ago
i only want to see golf-robot burying tiger woods at golf...
pedalla 8 months ago
This has been flagged as spam show
@pedalla he will retire before we see that kind of robot
singularku 8 months ago
But ... Will it blend ?
manemanemane11 8 months ago 5
@manemanemane11 LOL!!!
Thales512 8 months ago
This has been flagged as spam show
my enthusiasm decreased right in the begining of the video when i saw the robot has a ground plan view :-/
scottyfactor 8 months ago
my enthusiasm decreased right in the begining of the video when i saw the robot has a ground plan view :-/
scottyfactor 8 months ago
Humans can't have the ceiling view of the table, so it's not fair.
goranvu 8 months ago
Wow, just the projector aiding a human playing is AWESOME!
bankashvids 8 months ago
Science. It works, bitches.
TucoBenedictoPax 8 months ago
пусть попробует в русский поигарть
задача: мягкий скат свояка через весь стол на сильном боковике.
Mrtuzikpuzik 9 months ago
ok, some comments from the producer of the video:
- the stroke difficulty is only calculated based on the angular deviation of the white ball, meaning how much in degrees one can aim to the left or the right still being able to pocket the object ball. This measurement implicit includes the distance white ball - object ball, object ball - hole and the angle to pocket. No weighting of factors is necessary.
- cushion shots might be possible as soon as we get a better pool table ;-)
Thomas
nierth 9 months ago 2
Всё ХУЙНЯ он МЕДЛЕЕНО играет
garik0093 9 months ago
De carajo!!!! y para jugar carambolas de fantasia? :P
davidcruzvargas 9 months ago
2:17, ???
alebzibzi 9 months ago
@alebzibzi exactly, the blue ball is touched
mudfly3 9 months ago
Can it perform cusion and draw shots?
GTI8855 9 months ago
this robot plays better than me atleast....
luv2break 9 months ago
stellen kann er aber noch nicht so gut^^
LongBeachCaddilac 9 months ago
good
Gerrardsw 9 months ago
can't wait for a table tennis one
stfuplznoname 9 months ago
poland beware :D
spiritusinvacuo 9 months ago
2:15 It touched the blue ball! Man, he even knows how to cheat like a human.
HerrBismarck 9 months ago 2
Minnesota FAT32s.
RobTFirefly 9 months ago
Your robot gets lucky missing the 15 @ 1:37 and caroming it off what looks like the 2 (or is it the 4?) - LOL!! Fun little machine you've got here.
LeonFleisherFan 9 months ago
grt...!!! wud be eager to see how would be the future. keep it up guys...
altafnazir 9 months ago
what about trick shots?
yscreth55 9 months ago
it would be boring as hell waiting for this motherfucker's turn to take a shot
rezaletAK 9 months ago 2
ENGADGET brot me here :D
bigbad1000do7a 9 months ago
aperature science
daytripper22 9 months ago
lol @ foul 2:16
TheSanjMister 9 months ago
but, can it get me a beer?
JogBird 9 months ago
2:17 THE ROBOT TOUCHES THE BLUE BALL LOL FAIL FAIL FAIL
ENGLANDTHEGREATEST 9 months ago
You could sell the projection system for human players as a product. I'm sure the market for that is considerable.
guorbatschow 9 months ago
now thats cheating
jayorn 9 months ago
pool is boring. But the robot is awesome!
zairman 9 months ago
Comment removed
zairman 9 months ago
機械人也要休閒活動?當然不是,這是德國慕尼黑理工大學設計了一個會打桌球的機械人,
這機械人的獨特之處是打桌球時,會運用2隻手,這講求機械人雙手的合作性,瞄準桌球則運用天花上的相機。
機械人還能靈活地在桌球桌和旁,走來走去。機械人接收來自相機的數據,便會在電腦化成圖像數據,然後控制機械人的手和身體移動,做出完美擊球。更與真人桌球般,擊球前想過多個擊球可行性
s3063215 9 months ago
Man,even the machine is better then me at pool
PsychoCatIndustries 9 months ago
This is amazing pioneering, and to think it wont take many decades to refine this technology to the point it is aesthetically pleasing, and integrated as a function into humanoid robots without the need for an overhead camera. It would give as a companion to play pool against. However little we may win :P
mvenom101 9 months ago
WOWOWOWWOWOW
premierclip 9 months ago
Comment removed
lupemaxman 9 months ago
1:15
"decreases almost linearly with stroke difficulty."
-.-
And how do you measure stroke difficulty?
If it's by successful potting rate then ye, its going to be linear.
Isanion 9 months ago 25
@Isanion Im assuming it would be some combination of angle to the pocket, distance from pocket and if there any balls near the line to the pocket.
OwnageEngine 9 months ago
@OwnageEngine but those factors need to calibrated: what's more influential, distance or angle? In what ratio? How much is the difficulty adjusted by having a ball kissing the white?
That data needs to be gathered by experimentation.
If they gathered it by observing humans then they could possibly claim that it shows the robot's ability is comparable to that of a human's in regard to potting difficulty, but they'd need to present a lot of data comparing the robot to humans to make that claim.
Isanion 9 months ago
@Isanion Its arbitrary. The stroke difficulty would be related to how inaccurate parts of the robot are. If it has trouble aiming at balls at the other side of the table, but can hit close ones accurately then distance to the ball would have more weighting. So something like D = k(Distance to ball) + m(Distance to pot) + n(sin(angle to pocket)) where k,m and n give some indiacation of how bad the robot is at certain things.
OwnageEngine 9 months ago
@OwnageEngine
if they did do it the way you suggest then that makes the comment about the linear relation between difficulty and success even more pointless.
All it would indicate is that they did a good job at guessing what types of shots the robot would be best and worst at.
Isanion 9 months ago
@Isanion How is this pointless? Its exactly the same way you measure difficulity in anything humans do. If we arent good at something, its more difficult for us. Same thing happens with the robot.
OwnageEngine 9 months ago
@Isanion Not necessarily. =|
Haroids 9 months ago
@Isanion Lol, exactly! They defined "stroke difficulty" as successful potting rate and then just verified that with more shots!
nikapostol 9 months ago
@Isanion No, he has an algorithm to calculate the stroke difficulty for every shot. He was showing that the reuslts of that algorithm correlate linearly with the actual success rate, which is evidence that the algorithm works. He uses this algorithm during play so the robot can choose the easiest shot.
Elavid 9 months ago
@Isanion i guess it all depends on what kinda porno it watches
failblorg 9 months ago
@Isanion distance from the hole , and if u want other factors like the visibility of the target
T0R0RIKI 8 months ago
@Isanion The stroke difficulty is determined using the "allowed angular difficulty", which is basically the stroke's "error" angle that allows the object ball to be pocketed. The smaller the angle is, the easier the stroke will be.
It's all on their website and article.
nburo 6 months ago
Foul long before
tingsry 5 months ago
i can still beat it...
palfreys1 9 months ago
foul at 2:14 !!!
littlestworkshop 9 months ago 134
2:17 !!
MrDebauche 9 months ago
2:17 !!
MrDebauche 9 months ago
Yay a way to cheat at pool!
zigxig 9 months ago
wow thats just amazing
y4aser 9 months ago