CVETJun 19

Semi-Supervised Vision-Language-Action Model

arXiv:2606.214935.6
Predicted impact top 77% in CV · last 90 daysOriginality Incremental advance
AI Analysis

Reduces the need for costly action-labeled demonstrations in robot learning, enabling efficient VLA model adaptation with limited supervision.

SemiVLA adapts VLA models to new environments using only 10% action-labeled trajectories, achieving 89.0% average success on LIBERO, outperforming supervised LoRA by 8.0 points.

Vision-Language-Action (VLA) models enable robots to predict actions directly from visual observations and language instructions, but adapting them to new environments still depends on costly action-labeled demonstrations. To reduce this dependence, we study semi-supervised VLA adaptation under limited supervision signals, where only a small portion of trajectories contain robot actions and the remaining trajectories provide action-unlabeled vision-language observations. Unlike standard semi-supervised learning, the missing supervision is an embodied action signal that must be visually grounded, language-consistent, physically feasible, and temporally stable. To address this problem, we propose SemiVLA, a self-distilled teacher-student framework that learns from reliable pseudo-actions on unlabeled trajectories. SemiVLA introduces a VLA-specific reliability controller to assess vision-language alignment, action feasibility, and temporal transition consistency, and further updates the teacher through a Bottleneck-Projected Alignment Update to avoid noisy feedback contamination. With OpenVLA as the backbone, SemiVLA consistently improves multiple PEFT strategies across LIBERO and CALVIN. Under 10\% labeled trajectories, SemiVLA with Selective LoRA achieves 89.0\% average success on LIBERO, outperforming supervised LoRA by 8.0 points without extra inference cost.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes