Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

3D Diffusion Policy: Generalizable Visuomotor Policy Learning via Simple 3D Representations

About

Imitation learning provides an efficient way to teach robots dexterous skills; however, learning complex skills robustly and generalizablely usually consumes large amounts of human demonstrations. To tackle this challenging problem, we present 3D Diffusion Policy (DP3), a novel visual imitation learning approach that incorporates the power of 3D visual representations into diffusion policies, a class of conditional action generative models. The core design of DP3 is the utilization of a compact 3D visual representation, extracted from sparse point clouds with an efficient point encoder. In our experiments involving 72 simulation tasks, DP3 successfully handles most tasks with just 10 demonstrations and surpasses baselines with a 24.2% relative improvement. In 4 real robot tasks, DP3 demonstrates precise control with a high success rate of 85%, given only 40 demonstrations of each task, and shows excellent generalization abilities in diverse aspects, including space, viewpoint, appearance, and instance. Interestingly, in real robot experiments, DP3 rarely violates safety requirements, in contrast to baseline methods which frequently do, necessitating human intervention. Our extensive evaluation highlights the critical importance of 3D representations in real-world robot learning. Videos, code, and data are available on https://3d-diffusion-policy.github.io .

Yanjie Ze, Gu Zhang, Kangning Zhang, Chenyuan Hu, Muhan Wang, Huazhe Xu• 2024

Related benchmarks

TaskDatasetResultRank
Robotic ManipulationRoboTwin 2.0
Average Success Rate84
64
Robot ManipulationAdroit
Pen Task Score80
50
Robotic ManipulationRoboTwin 1.0
Success Rate81
48
Long-horizon robotic manipulationCalvin ABC->D
Task 1 Success Rate28.3
34
Robotic Tabletop ManipulationRoboCasa GR1 Tabletop Tasks
Average Success Rate33
28
Coffee Making/HandlingRobomimic MimicGen Coffee (D2)
Success Rate34
25
Robotic ManipulationRoboTwin 2.0 (test)
Average Success Rate77.8
22
Robot ManipulationMetaWorld 50 tasks
Success Rate (Easy)90.9
21
Robotic ManipulationAdroit and MetaWorld
Average Success Rate78.3
21
Bimanual ManipulationRLBench 2
Push Box Success Rate56
20
Showing 10 of 304 rows
...

Other info

Code

Follow for update