Guide-yakabatana mibvunzo · Pakati Nhanho
Vision-Language-Action Models yeRobotics Quiz
Dzidza kuti mamodheru echiono-mutauro-chiito anosanganisa sei kuisa kamera nemirayiridzo nemirairo yerobhoti, uye kuti humbowo hwawo hunobva kupi.
Mubvunzo 1 ye 8
Ndeapi marudzi matatu ekupinza/zvinobuda anotsanangura Vision-Language-Action (VLA) modhi?
Tarisa mubvunzo wese8 mibvunzo
Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.
- Ndeapi marudzi matatu ekupinza/zvinobuda anotsanangura Vision-Language-Action (VLA) modhi?
- Mifananidzo yekamera, dzidziso yemutauro, uye zviito zvemarobhoti
- Audio, zvinyorwa, uye mifananidzo
- Sensor logs, GPS, uye mamiriro ebhatiri
- Code, compiler inobuda, uye bvunzo mhinduro
- Ko Google DeepMind's RT-2 yakamiririra sei zviito zvemarobhoti kuti transformer igozvigadzira?
- Sezvinoenderera mberi zvinoyangarara-nzvimbo mavheta chete
- Sezvine hungwaru chiito tokeni, zvakafanana nemavara tokeni
- Sezvo ROS meseji XML
- Semibayiro mibairo yekusimbisa kudzidza
- Chii chinonzi 'emergent transfer' chinorevei mumamiriro ezvinhu eVLAs?
- Robhoti rinopisa uye rinodzima
- Marobhoti maviri anochinjanisa software yavo
- Webhu-chiyero ruzivo kubva pretraining inobatsira robhoti kubata zvinhu zvitsva uye mirairo
- Zviito zvinoverengerwa pasina chero neural network
- Chii chakakosha mukana wepi-0's diffusion/flow-matching action musoro uchienzaniswa neasing discrete action tokens?
- Inobvisa kudiwa kwekamera chero ipi zvayo
- Inoshanda chete mukufananidza
- Inobvisa dzidziso yemutauro
- Iyo inogadzira yakatsetseka, yakakwira-frequency chiito chunks kune mamwe dexterous control
- Sei iyo Yakavhurika X-Embodiment dataset ine basa kune VLAs?
- Inoputira data remarobhoti mumhando dzakawanda dzemarobhoti kugonesa kuyambuka-embodiment generalization
- Icho chiyereso chete chekutora mufananidzo
- Inochengeta data rerobhoti rimwe chete
- Iboka remavara emutauro chete
- Which data sources did DeepMind combine when co-fine-tuning RT-2?
- Only a fixed list of hand-written joint commands
- Robot demonstration data and web data with a pretrained vision-language model
- Only audio recordings without robot observations
- Only random action tokens with no examples
- What kinds of commands can an RT-2-style action representation encode?
- Only natural-language explanations with no robot controls
- A fixed video caption for each camera frame
- Discretized robot motion and gripper commands represented as tokens
- A robot’s hardware serial number
- How broad was the RT-2 evaluation of emergent capabilities reported by Google DeepMind?
- Exactly 12 trials
- One trial for every robot type worldwide
- No physical robot trials
- More than 6,000 robotic trials