Data-Asymmetric Latent Imagination and Reranking for 3D Robotic Imitation Learning
2026-05-11 • Robotics
Robotics
AI summaryⓘ
The authors address the problem that robotic learning usually expects perfect demonstrations, but real data often includes imperfect or failed attempts. Instead of ignoring this imperfect data, they created DALI-R, a system that learns from mixed-quality demonstrations by imagining possible futures using 3D point cloud data and by re-ranking potential actions to pick better ones. They tested their approach with two different types of 3D learning policies and found it improved success rates by about 7% without much extra computing cost. This way, robots can learn better from more realistic, imperfect data.
robotic imitation learning3D point cloudslatent world modeltask completion scorerdata-asymmetric learningdiffusion policiesflow matchingtrajectory rerankingAdroit benchmarkMetaWorld benchmark
Authors
Lianghao Luo, Xizhou Bu, Ruyan Liu, Qingqiu Huang, Chufeng Tang, Xiaoshuai Hao, Hongbo Wang, Wei Li
Abstract
Robotic imitation learning typically assumes access to optimal demonstrations, yet real-world data collection often yields suboptimal, exploratory, or even failed trajectories. Discarding such data wastes valuable information about environment dynamics and failure modes, which can instead be leveraged to improve decision-making. While 3D policies reduce reliance on high-quality demonstrations through strong spatial generalization, they still require large-scale data to achieve high task success. To address this, we propose DALI-R, a Data-Asymmetric Latent Imagination and Reranking framework for 3D robotic imitation learning from mixed-quality trajectories. It learns a Latent World Model over 3D point clouds for imagined rollouts and a Task Completion Scorer that reranks candidate action chunks, improving decision-making without additional high-quality demonstrations. We instantiate DALI-R with both diffusion and efficient flow-matching policies and evaluate it on Adroit and MetaWorld benchmarks. Across the two evaluated 3D base policies, DALI-R achieves an average $6.8$\% improvement in success rate while incurring less than $0.7\times$ additional inference overhead.