@xrobohub: 🎾🤖 INSANE! Researchers Can't Beat This #TennisRobot Anymore! #LATENT is Pro. #humanoid #robot #AI #RobotPower #EmbodiedAI #PhysicalAI #robotics

XRoboHub
XRoboHub
Open In TikTok:
Region: SG
Sunday 15 March 2026 16:44:59 GMT
264051
6158
263
1358

Music

Download

Comments

saidee888
saidee1798 :
This requires an insane amount of extreme computing speed..
2026-03-16 04:47:36
129
tshihoutang
tshihoutang_TONY :
researcher cannot beat it because he purposely feed it the ball..okay... so don't pretend .. don't do fake thing
2026-03-28 03:48:12
3
eisernefaust7
Eiserne faust :
Ai
2026-03-15 17:12:32
18
franken.fan8
Franken Fan :
No fooling☝️. this is still AI😁
2026-03-15 21:44:24
5
rainbowneil
haaaaaan :
さすがにAiだろ
2026-04-25 00:42:30
2
bingoddp6
Bingo :
Everyone who has ever played tennis will still beat it
2026-03-17 20:58:41
3
5sk1f
5k1f :
2026-03-15 19:15:01
47
userehj4x0nnvm
かばちん :
すごいな、ロボット技術が上がっていく過程を見るのは楽しい
2026-03-17 01:47:11
13
asmr_777_try
The ASMR Try :
imagine 1 year later. the human wont be able to beat it. it can practice 24/7 less charging time.
2026-04-18 00:50:54
6
user49372612180671
王峰 :
focus on the shadow,its real。
2026-03-16 14:45:51
7
up.in.there
Ricky Brown :
we are spending thousands of hours to train machines and not a minute to say hi to a neighbour or a stranger that we always see on a metro every morning in a train.
2026-04-22 13:14:06
1
sggibson
boy :
Key Details of the Video The Robot: A modified Unitree G1 humanoid robot equipped with a 3D-printed racket adapter and reflective markers for motion tracking. The Technology: Researchers from Tsinghua and Peking Universities developed the LATENT (Learning Athletic humanoid TEnnis skills from imperfect human motion daTa) framework. How it Learned: Instead of needing perfect data, the AI was trained on "imperfect" fragments of human motion (about 5 hours of data) and then "practiced" in a digital simulator before playing on a real court. Performance: In real-world tests, the robot successfully returned incoming balls with a 90% success rate for forehands and 78% for backhands.
2026-03-18 13:47:41
6
livedream.ai
LiveDream.ca :
Ai
2026-05-12 18:38:25
1
ahmaddc0
dhaniyono :
gak ikut lomba marathon??
2026-04-21 18:11:31
0
u.klow
u klow :
ai
2026-03-18 05:59:11
0
richspring7
GIAN :
AI 😒😒😡
2026-05-04 20:19:42
1
sggibson
boy :
Real or AI ?
2026-03-18 07:07:21
5
jp.nonono
ののの :
テスラのオプティマス?すごいね重心進行方向に倒して移動してる、しかも物持ってるしボールの衝撃も受けれてる
2026-03-27 08:50:17
1
user136152350569
バルカン :
まだこれは無理だろ!(笑)😅
2026-03-16 05:20:37
3
smosea
SMOSEA :
Tu imagines dans combien de temps des robots sur le circuit ATP
2026-03-16 08:34:42
1
myung.soo.han
myung.soo han :
ai
2026-03-17 06:54:49
0
klgan4
klgan :
If this is true, it demonstrates the achievement so far in robot tech. The movements in tennis or badminton are so random and fast, and for AI and robots to achieve this level now is really something.
2026-05-08 16:21:23
1
the.wealth.future
The Wealth Future :
How long until we have AI friends?
2026-03-16 08:20:41
3
caerukaixin
shizuka :
すごくない?練習すればプロになれるかもね👍
2026-03-25 04:57:56
2
To see more videos from user @xrobohub, please go to the Tikwm homepage.

Other Videos


About