Perfect Demo Makes Poor Teacher: Learning Robust Alignment from Critical Motion Segments

• URL: http://arxiv.org/abs/2606.15587v1
• Date: Sun, 14 Jun 2026 04:15:29 GMT
• Title: Perfect Demo Makes Poor Teacher: Learning Robust Alignment from Critical Motion Segments
• Abstract summary: Expert demonstrations are widely assumed to be the gold standard for robot imitation learning.<n>Yet for fine-grained manipulation such as insertion, stacking, and alignment, we uncover a counterintuitive failure mode: fluent demonstrations can be poor teachers.<n>We introduce STAIR, a compact dynamic feature that bridges the vision-language model and the action expert, distilling the short-horizon motion already recorded in each trajectory into dense, motion-aware supervision.
• Score: 53.85970147743299
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Expert demonstrations are widely assumed to be the gold standard for robot imitation learning. Yet for fine-grained manipulation such as insertion, stacking, and alignment, we uncover a counterintuitive failure mode: fluent demonstrations can be poor teachers. A skilled teleoperator compresses the decisive moments of alignment and recovery into a brief temporal window, leaving the policy flooded with redundant free-space motion and starved of supervision exactly where precision determines success. We address this bottleneck at two levels. At the data level, slowing down near alignment and resampling critical segments both help, yet the gain comes mainly from broadening the coverage of recovery states the policy must learn, not from reweighting frames it already has. Such data-side fixes, however, leave the policy's per-frame view untouched: a single image still maps directly to an action, and the local motion that governs correction stays implicit. We therefore turn to the representation level and introduce STAIR (\textbf{S}patio-\textbf{T}emporal feature \textbf{A}s an \textbf{I}nterface for \textbf{R}obot learning), a compact dynamic feature that bridges the vision-language model and the action expert, distilling the short-horizon motion already recorded in each trajectory into dense, motion-aware supervision. Trained on fluent data alone, STAIR recovers most of the deliberate-demonstration gain ($50.0$ to $62.2\%$ overall, approaching the $64.4\%$ of deliberate demonstrations). These results call for a more pedagogical view of robot data, optimized for machine learnability rather than human efficiency alone.

Related papers

• Learning to Move Before Learning to Do: Task-Agnostic pretraining for VLAs [46.61554174146852]
  Task-Agnostic Pretraining (TAP) is a two-stage framework that first learns transferable motor priors from cheap, unlabeled interaction data.<n>On the SIMPLER benchmark, TAP matches models trained on over 1M expert trajectories while using orders of magnitude less labeled data.<n>On a real-world WidowX platform, TAP retains 25% success under camera perturbations where internet-scale baselines collapse to 0%.
  arXiv  Detail & Related papers  (2026-07-02T17:33:37Z)
• Sensorimotor World Models: Perception for Action via Inverse Dynamics [59.121736411156384]
  We introduce a sensorimotor world model trained end-to-end with inverse dynamics regularization.<n>It prevents representation collapse and induces action-aligned representations.<n>It learns compact, interpretable latent spaces and enables competitive planning performance across simple 2D and 3D control tasks.
  arXiv  Detail & Related papers  (2026-06-18T11:25:16Z)
• AffordanceVLA: A Vision-Language-Action Model Empowering Action Generation through Affordance-Aware Understanding [55.535374902204296]
  AffordanceVLA is a unified framework that introduces structured affordance forecasting as a task-oriented intermediate representation.<n>We show that AffordanceVLA achieves strong performance across diverse manipulation scenarios.
  arXiv  Detail & Related papers  (2026-06-04T13:28:51Z)
• Partially Observable Adversarial Patch Attacks on Vision-Language-Action Models in Robotics [21.834006622805678]
  Vision-language-action (VLA) models are gaining attention in robotics, yet their robustness to adversarial attacks remains largely unexplored.<n>We formulate a partially observable threat model, where an adversary can exploit only a short prefix of the trajectory to generate a fixed patch applied to all subsequent frames.<n>First, we localize the patch using the model's attention maps to identify visually critical regions that correspond to the full instruction.<n>Then, we optimize the patch to disrupt the semantic grounding of target objects and increase the curvature of action trajectories.
  arXiv  Detail & Related papers  (2026-06-02T12:19:28Z)
• WM-DAgger: Enabling Efficient Data Aggregation for Imitation Learning with World Models [22.67666648771421]
  We propose WM-DAgger, an efficient data aggregation framework.<n>We focus on manipulation tasks with an eye-in-hand robotic arm and only few-shot demonstrations.<n>We extensively validate WM-DAgger on multiple real-world robotic tasks.
  arXiv  Detail & Related papers  (2026-04-13T11:49:11Z)
• Learning from Imperfect Demonstrations via Temporal Behavior Tree-Guided Trajectory Repair [5.4872967954342675]
  Learning robot control policies from demonstrations is a powerful paradigm, yet real-world data is often suboptimal, noisy, or otherwise imperfect.<n>We present a formal framework that leverages Temporal Behavior Trees (TBT) to repair suboptimal trajectories prior to their use in downstream policy learning.<n>We demonstrate the effectiveness of this framework on discrete grid-world navigation and continuous single and multi-agent reach-avoid tasks.
  arXiv  Detail & Related papers  (2026-04-05T18:49:45Z)
• CLAP: Contrastive Latent Action Pretraining for Learning Vision-Language-Action Models from Human Videos [73.51386721543135]
  We propose Contrastive Latent Action Pretraining (CLAP), a framework that aligns the visual latent space from videos with a proprioceptive latent space from robot trajectories.<n>CLAP maps video transitions onto a quantized, physically executable codebook.<n>We introduce a dual-formulation VLA framework offering both CLAP-NTP, an autoregressive model excelling at instruction following and object generalization, and CLAP-RF, a Rectified Flow-based policy designed for high-frequency, precise manipulation.
  arXiv  Detail & Related papers  (2026-01-07T16:26:33Z)
• Weakly and Self-Supervised Class-Agnostic Motion Prediction for Autonomous Driving [52.79390062794558]
  We investigate weakly and self-supervised class-agnostic motion prediction from LiDAR point clouds.<n>We propose a novel weakly supervised paradigm that replaces motion annotations with fully or partially annotated (1%, 0.1%) foreground/background masks for supervision.
  arXiv  Detail & Related papers  (2025-09-16T14:22:45Z)
• Self-Augmented Robot Trajectory: Efficient Imitation Learning via Safe Self-augmentation with Demonstrator-annotated Precision [2.3548641190233264]
  Self-Augmented Robot Trajectory (SART) is a framework that enables policy learning from a single human demonstration.<n>SART achieves substantially higher success rates than policies trained solely on human-collected demonstrations.
  arXiv  Detail & Related papers  (2025-09-11T23:10:56Z)
• Dexplore: Scalable Neural Control for Dexterous Manipulation from Reference-Scoped Exploration [58.4036440289082]
  Hand-object motion-capture (MoCap) offer large-scale, contact-rich demonstrations and hold promise for dexterous robotic scopes.<n>We introduce Dexplore, a unified single-loop optimization that performs repositories and tracking to learn robot control policies directly from MoCap at scale.
  arXiv  Detail & Related papers  (2025-09-11T17:59:07Z)
• MoSiC: Optimal-Transport Motion Trajectory for Dense Self-Supervised Learning [66.53533434848369]
  We propose a motion-guided self-learning framework that learns densely consistent representations.<n>We improve state-of-the-art by 1% to 6% on six image and video datasets and four evaluation benchmarks.
  arXiv  Detail & Related papers  (2025-06-10T11:20:32Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.