Learning Foresight without Explicit Trajectories for 3D Diffusion Policies

Hugging Face Daily Papers Papers
3d-diffusion-policies robotic-manipulation foresight-learning ai-research movement-trend-guidance

Summary

Introduces Movement Trend Guidance to enhance 3D diffusion policies in robotic manipulation by providing foresight without explicit trajectories, achieving improved performance on benchmarks like RoboTwin2.0 and LIBERO-40.

3D diffusion policies are strong at generating geometrically grounded actions from current observations, but successful manipulation requires not only knowing what motion is feasible now, but also anticipating where the interaction is heading. Existing policies largely leave such foresight to emerge implicitly from action learning. We introduce Movement Trend Guidance, a simple but effective way to provide this foresight without introducing an explicit plan. From a short observation history, the policy learns a compact latent representation of interaction evolution. During training, sparse future gripper states supervise this representation; at inference, only the latent is retained as future-oriented conditioning alongside the current observation. The latent provides global conditioning for action generation, while an additional gated FiLM branch is used only at the UNet bottleneck. Despite adding only 3.52% more parameters to DP3, our method preserves the original dense-action and receding-horizon formulation and consistently improves upon DP3 across RoboTwin2.0, LIBERO-40, and DexArt. It reaches 62.8% vs. 56.1% in 50-task RoboTwin2.0 mixed training, 71.93% vs. 37.08% on LIBERO-40, and 72.0% vs. 49.0% on five real-robot tasks. These results show that a diffusion policy can benefit substantially from knowing where an interaction is heading, without being told exactly where to move.
Original Article
View Cached Full Text

Cached at: 09/21/26, 03:21 AM

Paper page - Learning Foresight without Explicit Trajectories for 3D Diffusion Policies

Source: https://huggingface.co/papers/2609.20669

Abstract

3Ddiffusionpoliciesarestrongatgeneratinggeometricallygroundedactionsfromcurrentobservations,butsuccessfulmanipulationrequiresnotonlyknowingwhatmotionisfeasiblenow,butalsoanticipatingwheretheinteractionisheading.Existingpolicieslargelyleavesuchforesighttoemergeimplicitlyfromactionlearning.WeintroduceMovementTrendGuidance,asimplebuteffectivewaytoprovidethisforesightwithoutintroducinganexplicitplan.Fromashortobservationhistory,thepolicylearnsacompactlatentrepresentationofinteractionevolution.Duringtraining,sparsefuturegripperstatessupervisethisrepresentation;atinference,onlythelatentisretainedasfuture-orientedconditioningalongsidethecurrentobservation.Thelatentprovidesglobalconditioningforactiongeneration,whileanadditionalgatedFiLMbranchisusedonlyattheUNetbottleneck.Despiteaddingonly3.52%moreparameterstoDP3,ourmethodpreservestheoriginaldense-actionandreceding-horizonformulationandconsistentlyimprovesuponDP3acrossRoboTwin2.0,LIBERO-40,andDexArt.Itreaches62.8%vs.56.1%in50-taskRoboTwin2.0mixedtraining,71.93%vs.37.08%onLIBERO-40,and72.0%vs.49.0%onfivereal-robottasks.Theseresultsshowthatadiffusionpolicycanbenefitsubstantiallyfromknowingwhereaninteractionisheading,withoutbeingtoldexactlywheretomove.

View arXiv pageView PDFProject pageGitHub0Add to collection

Get this paper in your agent:

hf papers read 2609\.20669

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2609.20669 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2609.20669 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2609.20669 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

From Noise to Control: Parameterized Diffusion Policies

arXiv cs.AI

This paper introduces Parameterized Diffusion Policy (PDP), a framework that makes diffusion policies controllable by conditioning on low-dimensional latent parameters, enabling smooth behavior interpolation and adaptation without retraining. It demonstrates improved performance on complex multimodal robot tasks in simulation and real-world experiments.

Anticipatory Reinforcement Learning for Trajectory Tracking

arXiv cs.LG

This paper introduces a predictive formulation for deep reinforcement learning that augments the state space with future reference horizons to enable anticipatory control for trajectory tracking. Simulation results show significant error reduction, though zero-shot transfer to physical hardware reveals a sim-to-real gap.